Datasets:

CodeGoat24
/

VideoDPO

Dataset card Data Studio Files Files and versions Community

Split (1)

train · 10k rows

Subsets and Splits

No community queries yet

The top public SQL queries from the community will appear here once available.

prompt stringlengths 4 8.37k	conversations listlengths 2 2	images sequencelengths 16 16
yellow snake with 2 long teeth, ultra realistic, moving, FHD, 8k -fps 24 -ar 9:16.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [yellow snake with 2 long teeth, ultra realistic, moving, FHD, 8k -fps 24 -ar 9:16.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/000629/0.png", "losevideos/000629/1.png", "losevideos/000629/2.png", "losevideos/000629/3.png", "losevideos/000629/4.png", "losevideos/000629/5.png", "losevideos/000629/6.png", "losevideos/000629/7.png", "winvideos/000629/0.png", "winvideos/000629/1.png", "winvideos/000629/2.png", "winvideos/000629/3.png", "winvideos/000629/4.png", "winvideos/000629/5.png", "winvideos/000629/6.png", "winvideos/000629/7.png" ]
show a tiger eating a cow.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [show a tiger eating a cow.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/008842/0.png", "losevideos/008842/1.png", "losevideos/008842/2.png", "losevideos/008842/3.png", "losevideos/008842/4.png", "losevideos/008842/5.png", "losevideos/008842/6.png", "losevideos/008842/7.png", "winvideos/008842/0.png", "winvideos/008842/1.png", "winvideos/008842/2.png", "winvideos/008842/3.png", "winvideos/008842/4.png", "winvideos/008842/5.png", "winvideos/008842/6.png", "winvideos/008842/7.png" ]
angry person with wide ears and funny look having a conversation.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [angry person with wide ears and funny look having a conversation.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/002025/0.png", "winvideos/002025/1.png", "winvideos/002025/2.png", "winvideos/002025/3.png", "winvideos/002025/4.png", "winvideos/002025/5.png", "winvideos/002025/6.png", "winvideos/002025/7.png", "losevideos/002025/0.png", "losevideos/002025/1.png", "losevideos/002025/2.png", "losevideos/002025/3.png", "losevideos/002025/4.png", "losevideos/002025/5.png", "losevideos/002025/6.png", "losevideos/002025/7.png" ]
A group of male friends gathers around a table to play cards during Diwali. One friend proudly shows his hand, which consists of three Kings in a game of Teen Patti, securing a winning move.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [A group of male friends gathers around a table to play cards during Diwali. One friend proudly shows his hand, which consists of three Kings in a game of Teen Patti, securing a winning move.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/001489/0.png", "winvideos/001489/1.png", "winvideos/001489/2.png", "winvideos/001489/3.png", "winvideos/001489/4.png", "winvideos/001489/5.png", "winvideos/001489/6.png", "winvideos/001489/7.png", "losevideos/001489/0.png", "losevideos/001489/1.png", "losevideos/001489/2.png", "losevideos/001489/3.png", "losevideos/001489/4.png", "losevideos/001489/5.png", "losevideos/001489/6.png", "losevideos/001489/7.png" ]
32k resolution, sharp, captured by Phantom High-Speed Camera hyper realistic whimsical dark fantasy atmosphere, vibrant colors, holographic butterflies flying to the camera, soap bubbles spinning around, dark forest, dawn.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [32k resolution, sharp, captured by Phantom High-Speed Camera hyper realistic whimsical dark fantasy atmosphere, vibrant colors, holographic butterflies flying to the camera, soap bubbles spinning around, dark forest, dawn.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/000116/0.png", "winvideos/000116/1.png", "winvideos/000116/2.png", "winvideos/000116/3.png", "winvideos/000116/4.png", "winvideos/000116/5.png", "winvideos/000116/6.png", "winvideos/000116/7.png", "losevideos/000116/0.png", "losevideos/000116/1.png", "losevideos/000116/2.png", "losevideos/000116/3.png", "losevideos/000116/4.png", "losevideos/000116/5.png", "losevideos/000116/6.png", "losevideos/000116/7.png" ]
metal females wearing black dresses walking -Full shot -neg close up.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [metal females wearing black dresses walking -Full shot -neg close up.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/008764/0.png", "losevideos/008764/1.png", "losevideos/008764/2.png", "losevideos/008764/3.png", "losevideos/008764/4.png", "losevideos/008764/5.png", "losevideos/008764/6.png", "losevideos/008764/7.png", "winvideos/008764/0.png", "winvideos/008764/1.png", "winvideos/008764/2.png", "winvideos/008764/3.png", "winvideos/008764/4.png", "winvideos/008764/5.png", "winvideos/008764/6.png", "winvideos/008764/7.png" ]
illustration of jessy lanza mecha by satoshi kon, horse, mecha, los angeles -motion 4.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [illustration of jessy lanza mecha by satoshi kon, horse, mecha, los angeles -motion 4.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/009790/0.png", "losevideos/009790/1.png", "losevideos/009790/2.png", "losevideos/009790/3.png", "losevideos/009790/4.png", "losevideos/009790/5.png", "losevideos/009790/6.png", "losevideos/009790/7.png", "winvideos/009790/0.png", "winvideos/009790/1.png", "winvideos/009790/2.png", "winvideos/009790/3.png", "winvideos/009790/4.png", "winvideos/009790/5.png", "winvideos/009790/6.png", "winvideos/009790/7.png" ]
: gentle wind, eye slowly twitching, hair flutters, full HD, 8k -camera zoom out.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [: gentle wind, eye slowly twitching, hair flutters, full HD, 8k -camera zoom out.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/004417/0.png", "losevideos/004417/1.png", "losevideos/004417/2.png", "losevideos/004417/3.png", "losevideos/004417/4.png", "losevideos/004417/5.png", "losevideos/004417/6.png", "losevideos/004417/7.png", "winvideos/004417/0.png", "winvideos/004417/1.png", "winvideos/004417/2.png", "winvideos/004417/3.png", "winvideos/004417/4.png", "winvideos/004417/5.png", "winvideos/004417/6.png", "winvideos/004417/7.png" ]
make a video of a happy shop keeper handing over the package to the customer.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [make a video of a happy shop keeper handing over the package to the customer.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/009599/0.png", "losevideos/009599/1.png", "losevideos/009599/2.png", "losevideos/009599/3.png", "losevideos/009599/4.png", "losevideos/009599/5.png", "losevideos/009599/6.png", "losevideos/009599/7.png", "winvideos/009599/0.png", "winvideos/009599/1.png", "winvideos/009599/2.png", "winvideos/009599/3.png", "winvideos/009599/4.png", "winvideos/009599/5.png", "winvideos/009599/6.png", "winvideos/009599/7.png" ]
An image of girl handing the magical seed to boy.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [An image of girl handing the magical seed to boy.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/001429/0.png", "losevideos/001429/1.png", "losevideos/001429/2.png", "losevideos/001429/3.png", "losevideos/001429/4.png", "losevideos/001429/5.png", "losevideos/001429/6.png", "losevideos/001429/7.png", "winvideos/001429/0.png", "winvideos/001429/1.png", "winvideos/001429/2.png", "winvideos/001429/3.png", "winvideos/001429/4.png", "winvideos/001429/5.png", "winvideos/001429/6.png", "winvideos/001429/7.png" ]
a swarm of UFOs gathering over a lush forest with a monument —ar9:16 f1.8 -fps24.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [a swarm of UFOs gathering over a lush forest with a monument —ar9:16 f1.8 -fps24.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/000168/0.png", "losevideos/000168/1.png", "losevideos/000168/2.png", "losevideos/000168/3.png", "losevideos/000168/4.png", "losevideos/000168/5.png", "losevideos/000168/6.png", "losevideos/000168/7.png", "winvideos/000168/0.png", "winvideos/000168/1.png", "winvideos/000168/2.png", "winvideos/000168/3.png", "winvideos/000168/4.png", "winvideos/000168/5.png", "winvideos/000168/6.png", "winvideos/000168/7.png" ]
bart simpson drink heineken in rio de janeiro.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [bart simpson drink heineken in rio de janeiro.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/004537/0.png", "losevideos/004537/1.png", "losevideos/004537/2.png", "losevideos/004537/3.png", "losevideos/004537/4.png", "losevideos/004537/5.png", "losevideos/004537/6.png", "losevideos/004537/7.png", "winvideos/004537/0.png", "winvideos/004537/1.png", "winvideos/004537/2.png", "winvideos/004537/3.png", "winvideos/004537/4.png", "winvideos/004537/5.png", "winvideos/004537/6.png", "winvideos/004537/7.png" ]
a paper boat floats on the water.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [a paper boat floats on the water.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/005048/0.png", "winvideos/005048/1.png", "winvideos/005048/2.png", "winvideos/005048/3.png", "winvideos/005048/4.png", "winvideos/005048/5.png", "winvideos/005048/6.png", "winvideos/005048/7.png", "losevideos/005048/0.png", "losevideos/005048/1.png", "losevideos/005048/2.png", "losevideos/005048/3.png", "losevideos/005048/4.png", "losevideos/005048/5.png", "losevideos/005048/6.png", "losevideos/005048/7.png" ]
cute baby version of darth maul.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [cute baby version of darth maul.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/007914/0.png", "losevideos/007914/1.png", "losevideos/007914/2.png", "losevideos/007914/3.png", "losevideos/007914/4.png", "losevideos/007914/5.png", "losevideos/007914/6.png", "losevideos/007914/7.png", "winvideos/007914/0.png", "winvideos/007914/1.png", "winvideos/007914/2.png", "winvideos/007914/3.png", "winvideos/007914/4.png", "winvideos/007914/5.png", "winvideos/007914/6.png", "winvideos/007914/7.png" ]
princess throwing playing ball higher, disney, 8k video.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [princess throwing playing ball higher, disney, 8k video.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/006850/0.png", "winvideos/006850/1.png", "winvideos/006850/2.png", "winvideos/006850/3.png", "winvideos/006850/4.png", "winvideos/006850/5.png", "winvideos/006850/6.png", "winvideos/006850/7.png", "losevideos/006850/0.png", "losevideos/006850/1.png", "losevideos/006850/2.png", "losevideos/006850/3.png", "losevideos/006850/4.png", "losevideos/006850/5.png", "losevideos/006850/6.png", "losevideos/006850/7.png" ]
apples fall from the tree-camera zoom out.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [apples fall from the tree-camera zoom out.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/003543/0.png", "winvideos/003543/1.png", "winvideos/003543/2.png", "winvideos/003543/3.png", "winvideos/003543/4.png", "winvideos/003543/5.png", "winvideos/003543/6.png", "winvideos/003543/7.png", "losevideos/003543/0.png", "losevideos/003543/1.png", "losevideos/003543/2.png", "losevideos/003543/3.png", "losevideos/003543/4.png", "losevideos/003543/5.png", "losevideos/003543/6.png", "losevideos/003543/7.png" ]
human girl ring smartphone.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [human girl ring smartphone.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/009398/0.png", "losevideos/009398/1.png", "losevideos/009398/2.png", "losevideos/009398/3.png", "losevideos/009398/4.png", "losevideos/009398/5.png", "losevideos/009398/6.png", "losevideos/009398/7.png", "winvideos/009398/0.png", "winvideos/009398/1.png", "winvideos/009398/2.png", "winvideos/009398/3.png", "winvideos/009398/4.png", "winvideos/009398/5.png", "winvideos/009398/6.png", "winvideos/009398/7.png" ]
The guy with his pants in the subway. Korean style.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [The guy with his pants in the subway. Korean style.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/001552/0.png", "winvideos/001552/1.png", "winvideos/001552/2.png", "winvideos/001552/3.png", "winvideos/001552/4.png", "winvideos/001552/5.png", "winvideos/001552/6.png", "winvideos/001552/7.png", "losevideos/001552/0.png", "losevideos/001552/1.png", "losevideos/001552/2.png", "losevideos/001552/3.png", "losevideos/001552/4.png", "losevideos/001552/5.png", "losevideos/001552/6.png", "losevideos/001552/7.png" ]
Ethan reaches out to the silver-furred Frostfang pup to pet it.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [Ethan reaches out to the silver-furred Frostfang pup to pet it.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/005955/0.png", "winvideos/005955/1.png", "winvideos/005955/2.png", "winvideos/005955/3.png", "winvideos/005955/4.png", "winvideos/005955/5.png", "winvideos/005955/6.png", "winvideos/005955/7.png", "losevideos/005955/0.png", "losevideos/005955/1.png", "losevideos/005955/2.png", "losevideos/005955/3.png", "losevideos/005955/4.png", "losevideos/005955/5.png", "losevideos/005955/6.png", "losevideos/005955/7.png" ]
epic landscape with dynamic clouds.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [epic landscape with dynamic clouds.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/004118/0.png", "losevideos/004118/1.png", "losevideos/004118/2.png", "losevideos/004118/3.png", "losevideos/004118/4.png", "losevideos/004118/5.png", "losevideos/004118/6.png", "losevideos/004118/7.png", "winvideos/004118/0.png", "winvideos/004118/1.png", "winvideos/004118/2.png", "winvideos/004118/3.png", "winvideos/004118/4.png", "winvideos/004118/5.png", "winvideos/004118/6.png", "winvideos/004118/7.png" ]
real life, Eagle flying film.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [real life, Eagle flying film.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/000786/0.png", "winvideos/000786/1.png", "winvideos/000786/2.png", "winvideos/000786/3.png", "winvideos/000786/4.png", "winvideos/000786/5.png", "winvideos/000786/6.png", "winvideos/000786/7.png", "losevideos/000786/0.png", "losevideos/000786/1.png", "losevideos/000786/2.png", "losevideos/000786/3.png", "losevideos/000786/4.png", "losevideos/000786/5.png", "losevideos/000786/6.png", "losevideos/000786/7.png" ]
square grid sci-fi 80's film style -neg "morphing, erratic fluctuation in motion, noisy, bad quality, distorted, poorly drawn, blurry, grainy, low resolution, oversaturated, lack of detail, inconsistent lighting bad hands, wrong hands, simplify, separated_body_parts, bad bodies, wrong bodies, blurry, distorted, low resolution, poorly drawing, poorly textured, poorly resolution" -motion 4 Message: 1 Attachment.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [square grid sci-fi 80's film style -neg "morphing, erratic fluctuation in motion, noisy, bad quality, distorted, poorly drawn, blurry, grainy, low resolution, oversaturated, lack of detail, inconsistent lighting bad hands, wrong hands, simplify, separated_body_parts, bad bodies, wrong bodies, blurry, distorted, low resolution, poorly drawing, poorly textured, poorly resolution" -motion 4 Message: 1 Attachment.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/001256/0.png", "winvideos/001256/1.png", "winvideos/001256/2.png", "winvideos/001256/3.png", "winvideos/001256/4.png", "winvideos/001256/5.png", "winvideos/001256/6.png", "winvideos/001256/7.png", "losevideos/001256/0.png", "losevideos/001256/1.png", "losevideos/001256/2.png", "losevideos/001256/3.png", "losevideos/001256/4.png", "losevideos/001256/5.png", "losevideos/001256/6.png", "losevideos/001256/7.png" ]
* Aaaaaa -neg no distortion, malformations of body parts or face no melding of characters, no morphing, -seed 000000009.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [* Aaaaaa -neg no distortion, malformations of body parts or face no melding of characters, no morphing, \n-seed 000000009.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/003744/0.png", "losevideos/003744/1.png", "losevideos/003744/2.png", "losevideos/003744/3.png", "losevideos/003744/4.png", "losevideos/003744/5.png", "losevideos/003744/6.png", "losevideos/003744/7.png", "winvideos/003744/0.png", "winvideos/003744/1.png", "winvideos/003744/2.png", "winvideos/003744/3.png", "winvideos/003744/4.png", "winvideos/003744/5.png", "winvideos/003744/6.png", "winvideos/003744/7.png" ]
ancient Greece, Kyros, an athletic sailor with determined eyes, carefully checks the women of his majestic wooden ship vector art ready, vivid colors, high details -GS5, Motion2.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [ancient Greece, Kyros, an athletic sailor with determined eyes, carefully checks the women of his majestic wooden ship vector art ready, vivid colors, high details -GS5, Motion2.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/005871/0.png", "winvideos/005871/1.png", "winvideos/005871/2.png", "winvideos/005871/3.png", "winvideos/005871/4.png", "winvideos/005871/5.png", "winvideos/005871/6.png", "winvideos/005871/7.png", "losevideos/005871/0.png", "losevideos/005871/1.png", "losevideos/005871/2.png", "losevideos/005871/3.png", "losevideos/005871/4.png", "losevideos/005871/5.png", "losevideos/005871/6.png", "losevideos/005871/7.png" ]
applique made of torn and compressed paper of different textures Message: Keep Smiling (Font: MODERN).	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [applique made of torn and compressed paper of different textures Message: Keep Smiling (Font: MODERN).]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/000626/0.png", "winvideos/000626/1.png", "winvideos/000626/2.png", "winvideos/000626/3.png", "winvideos/000626/4.png", "winvideos/000626/5.png", "winvideos/000626/6.png", "winvideos/000626/7.png", "losevideos/000626/0.png", "losevideos/000626/1.png", "losevideos/000626/2.png", "losevideos/000626/3.png", "losevideos/000626/4.png", "losevideos/000626/5.png", "losevideos/000626/6.png", "losevideos/000626/7.png" ]
man goes office business man.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [man goes office business man.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/004045/0.png", "losevideos/004045/1.png", "losevideos/004045/2.png", "losevideos/004045/3.png", "losevideos/004045/4.png", "losevideos/004045/5.png", "losevideos/004045/6.png", "losevideos/004045/7.png", "winvideos/004045/0.png", "winvideos/004045/1.png", "winvideos/004045/2.png", "winvideos/004045/3.png", "winvideos/004045/4.png", "winvideos/004045/5.png", "winvideos/004045/6.png", "winvideos/004045/7.png" ]
rainy window, dark restaurant, 1970s -ar 4:3 -motion 3.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [rainy window, dark restaurant, 1970s -ar 4:3 -motion 3.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/009749/0.png", "winvideos/009749/1.png", "winvideos/009749/2.png", "winvideos/009749/3.png", "winvideos/009749/4.png", "winvideos/009749/5.png", "winvideos/009749/6.png", "winvideos/009749/7.png", "losevideos/009749/0.png", "losevideos/009749/1.png", "losevideos/009749/2.png", "losevideos/009749/3.png", "losevideos/009749/4.png", "losevideos/009749/5.png", "losevideos/009749/6.png", "losevideos/009749/7.png" ]
princess, close-up, outdoor, crowd watching, cinematography, HD, 8k -fps 24 -gs 12 -motion 2 -camera zoom out -neg eformed, bad anatomy, disfigured, poorly drawn face, mutation, mutated, extra limb, ugly, disgusting, poorly drawn hands, missing limb, floating limbs, disconnected limbs, malformed hands, blurry.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [princess, \nclose-up, outdoor, crowd watching, cinematography, HD, 8k -fps 24 -gs 12 -motion 2 -camera zoom out -neg eformed, bad anatomy, disfigured, poorly drawn face, mutation, mutated, extra limb, ugly, disgusting, poorly drawn hands, missing limb, floating limbs, disconnected limbs, malformed hands, blurry.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/000792/0.png", "winvideos/000792/1.png", "winvideos/000792/2.png", "winvideos/000792/3.png", "winvideos/000792/4.png", "winvideos/000792/5.png", "winvideos/000792/6.png", "winvideos/000792/7.png", "losevideos/000792/0.png", "losevideos/000792/1.png", "losevideos/000792/2.png", "losevideos/000792/3.png", "losevideos/000792/4.png", "losevideos/000792/5.png", "losevideos/000792/6.png", "losevideos/000792/7.png" ]
robot walking in mars.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [robot walking in mars.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/008338/0.png", "losevideos/008338/1.png", "losevideos/008338/2.png", "losevideos/008338/3.png", "losevideos/008338/4.png", "losevideos/008338/5.png", "losevideos/008338/6.png", "losevideos/008338/7.png", "winvideos/008338/0.png", "winvideos/008338/1.png", "winvideos/008338/2.png", "winvideos/008338/3.png", "winvideos/008338/4.png", "winvideos/008338/5.png", "winvideos/008338/6.png", "winvideos/008338/7.png" ]
A realistc high definition beautiful representation of a muscular blonde man we see him from the back he is in a park and we can see him running fast --ar 9:16.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [A realistc high definition beautiful representation of a muscular blonde man we see him from the back he is in a park and we can see him running fast --ar 9:16.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/007531/0.png", "winvideos/007531/1.png", "winvideos/007531/2.png", "winvideos/007531/3.png", "winvideos/007531/4.png", "winvideos/007531/5.png", "winvideos/007531/6.png", "winvideos/007531/7.png", "losevideos/007531/0.png", "losevideos/007531/1.png", "losevideos/007531/2.png", "losevideos/007531/3.png", "losevideos/007531/4.png", "losevideos/007531/5.png", "losevideos/007531/6.png", "losevideos/007531/7.png" ]
Crayon Shin chan and Doraemon compete with their hands to burn a copper gong.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [Crayon Shin chan and Doraemon compete with their hands to burn a copper gong.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/005121/0.png", "winvideos/005121/1.png", "winvideos/005121/2.png", "winvideos/005121/3.png", "winvideos/005121/4.png", "winvideos/005121/5.png", "winvideos/005121/6.png", "winvideos/005121/7.png", "losevideos/005121/0.png", "losevideos/005121/1.png", "losevideos/005121/2.png", "losevideos/005121/3.png", "losevideos/005121/4.png", "losevideos/005121/5.png", "losevideos/005121/6.png", "losevideos/005121/7.png" ]
Villagers honored 18 year old boy --ar16:9.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [Villagers honored 18 year old boy --ar16:9.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/007833/0.png", "winvideos/007833/1.png", "winvideos/007833/2.png", "winvideos/007833/3.png", "winvideos/007833/4.png", "winvideos/007833/5.png", "winvideos/007833/6.png", "winvideos/007833/7.png", "losevideos/007833/0.png", "losevideos/007833/1.png", "losevideos/007833/2.png", "losevideos/007833/3.png", "losevideos/007833/4.png", "losevideos/007833/5.png", "losevideos/007833/6.png", "losevideos/007833/7.png" ]
absract liquid background, 16:9 Video.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [absract liquid background, 16:9 Video.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/009907/0.png", "winvideos/009907/1.png", "winvideos/009907/2.png", "winvideos/009907/3.png", "winvideos/009907/4.png", "winvideos/009907/5.png", "winvideos/009907/6.png", "winvideos/009907/7.png", "losevideos/009907/0.png", "losevideos/009907/1.png", "losevideos/009907/2.png", "losevideos/009907/3.png", "losevideos/009907/4.png", "losevideos/009907/5.png", "losevideos/009907/6.png", "losevideos/009907/7.png" ]
a woman in a dress posing for a picture.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [a woman in a dress posing for a picture.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/004062/0.png", "winvideos/004062/1.png", "winvideos/004062/2.png", "winvideos/004062/3.png", "winvideos/004062/4.png", "winvideos/004062/5.png", "winvideos/004062/6.png", "winvideos/004062/7.png", "losevideos/004062/0.png", "losevideos/004062/1.png", "losevideos/004062/2.png", "losevideos/004062/3.png", "losevideos/004062/4.png", "losevideos/004062/5.png", "losevideos/004062/6.png", "losevideos/004062/7.png" ]
HEKA the magicians is a witcher living in the the egyptian ancient civilisation city of Zail , dressing in cool advanced royal vultures , 4k, cinematic.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [HEKA the magicians is a witcher living in the the egyptian ancient civilisation city of Zail , dressing in cool advanced royal vultures , 4k, cinematic.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/003468/0.png", "winvideos/003468/1.png", "winvideos/003468/2.png", "winvideos/003468/3.png", "winvideos/003468/4.png", "winvideos/003468/5.png", "winvideos/003468/6.png", "winvideos/003468/7.png", "losevideos/003468/0.png", "losevideos/003468/1.png", "losevideos/003468/2.png", "losevideos/003468/3.png", "losevideos/003468/4.png", "losevideos/003468/5.png", "losevideos/003468/6.png", "losevideos/003468/7.png" ]
Oprah pops a pimple.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [Oprah pops a pimple.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/004583/0.png", "losevideos/004583/1.png", "losevideos/004583/2.png", "losevideos/004583/3.png", "losevideos/004583/4.png", "losevideos/004583/5.png", "losevideos/004583/6.png", "losevideos/004583/7.png", "winvideos/004583/0.png", "winvideos/004583/1.png", "winvideos/004583/2.png", "winvideos/004583/3.png", "winvideos/004583/4.png", "winvideos/004583/5.png", "winvideos/004583/6.png", "winvideos/004583/7.png" ]
women tied in dark godown with face covered lying on the floor, dark cinematic -camera zoom out -neg morphing, erratic fluctuation in motion, noisy, bad quality, distorted, poorly drawn, blurry, grainy, low resolution, oversaturated, lack of detail, inconsistent lighting, 16:9.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [women tied in dark godown with face covered lying on the floor, dark cinematic -camera zoom out -neg morphing, erratic fluctuation in motion, noisy, bad quality, distorted, poorly drawn, blurry, grainy, low resolution, oversaturated, lack of detail, inconsistent lighting, 16:9.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/007397/0.png", "winvideos/007397/1.png", "winvideos/007397/2.png", "winvideos/007397/3.png", "winvideos/007397/4.png", "winvideos/007397/5.png", "winvideos/007397/6.png", "winvideos/007397/7.png", "losevideos/007397/0.png", "losevideos/007397/1.png", "losevideos/007397/2.png", "losevideos/007397/3.png", "losevideos/007397/4.png", "losevideos/007397/5.png", "losevideos/007397/6.png", "losevideos/007397/7.png" ]
a man with a hat stands in front of the camera, black and white picture, man smoking, angular, raining.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [a man with a hat stands in front of the camera, black and white picture, man smoking, angular, raining.]" }, { "from": "gpt", "value": "Video 2 is better than Video 1." } ]	[ "losevideos/004362/0.png", "losevideos/004362/1.png", "losevideos/004362/2.png", "losevideos/004362/3.png", "losevideos/004362/4.png", "losevideos/004362/5.png", "losevideos/004362/6.png", "losevideos/004362/7.png", "winvideos/004362/0.png", "winvideos/004362/1.png", "winvideos/004362/2.png", "winvideos/004362/3.png", "winvideos/004362/4.png", "winvideos/004362/5.png", "winvideos/004362/6.png", "winvideos/004362/7.png" ]
a woman being chased by babies, in the style of professional cinematography, feature film -gs 35 -motion 4 -camera pan left.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [a woman being chased by babies, in the style of professional cinematography, feature film -gs 35 -motion 4 -camera pan left.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/004711/0.png", "winvideos/004711/1.png", "winvideos/004711/2.png", "winvideos/004711/3.png", "winvideos/004711/4.png", "winvideos/004711/5.png", "winvideos/004711/6.png", "winvideos/004711/7.png", "losevideos/004711/0.png", "losevideos/004711/1.png", "losevideos/004711/2.png", "losevideos/004711/3.png", "losevideos/004711/4.png", "losevideos/004711/5.png", "losevideos/004711/6.png", "losevideos/004711/7.png" ]
oversized dark red colored Devil stands still spreading his wings. Three terrified looking people crawling under his wings.	[ { "from": "human", "value": "<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n <image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\n<image>\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos. You are given a text caption and the frames of two generated videos based on that caption. Your task is to evaluate and compare two videos based on two key criteria:\n1. Alignment with the Caption: Assess how well each video aligns with the provided caption. Consider the accuracy of depicted objects, their relationships, and attributes as described in the caption.\n2. Overall Video Quality: Examine the visual quality of each video, including clarity, detail preservation, color accuracy, and overall aesthetic appeal.\nCompare both videos using the above criteria and select the one that better aligns with the caption while exhibiting superior visual quality.\nProvide a clear conclusion such as \"Video 1 is better than Video 2.\", \"Video 2 is better than Video 1.\" and \"Both videos are equally good.\"\nYour task is provided as follows:\nText Caption: [oversized dark red colored Devil stands still spreading his wings. Three terrified looking people crawling under his wings.]" }, { "from": "gpt", "value": "Video 1 is better than Video 2." } ]	[ "winvideos/006165/0.png", "winvideos/006165/1.png", "winvideos/006165/2.png", "winvideos/006165/3.png", "winvideos/006165/4.png", "winvideos/006165/5.png", "winvideos/006165/6.png", "winvideos/006165/7.png", "losevideos/006165/0.png", "losevideos/006165/1.png", "losevideos/006165/2.png", "losevideos/006165/3.png", "losevideos/006165/4.png", "losevideos/006165/5.png", "losevideos/006165/6.png", "losevideos/006165/7.png" ]