Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -1,3 +1,156 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
from fastapi import FastAPI, UploadFile, File
|
| 2 |
import cv2
|
| 3 |
import torch
|
|
@@ -118,14 +271,25 @@ def convertir_sequences_en_json(dataframe):
|
|
| 118 |
events.append(event)
|
| 119 |
return events
|
| 120 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 121 |
@app.post("/analyze_video/")
|
| 122 |
async def analyze_video(file: UploadFile = File(...)):
|
| 123 |
-
|
| 124 |
-
|
|
|
|
|
|
|
| 125 |
|
| 126 |
-
|
|
|
|
| 127 |
json_result = convertir_sequences_en_json(dataframe_sequences)
|
| 128 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 129 |
|
| 130 |
@app.get("/", response_class=HTMLResponse)
|
| 131 |
async def index():
|
|
|
|
| 1 |
+
# from fastapi import FastAPI, UploadFile, File
|
| 2 |
+
# import cv2
|
| 3 |
+
# import torch
|
| 4 |
+
# import pandas as pd
|
| 5 |
+
# from PIL import Image
|
| 6 |
+
# from transformers import AutoImageProcessor, AutoModelForImageClassification
|
| 7 |
+
# from tqdm import tqdm
|
| 8 |
+
# import json
|
| 9 |
+
# import shutil
|
| 10 |
+
# from fastapi.middleware.cors import CORSMiddleware
|
| 11 |
+
# from fastapi.responses import HTMLResponse
|
| 12 |
+
|
| 13 |
+
# app = FastAPI()
|
| 14 |
+
|
| 15 |
+
# # Add CORS middleware to allow requests from localhost:8080 (or any origin you specify)
|
| 16 |
+
# app.add_middleware(
|
| 17 |
+
# CORSMiddleware,
|
| 18 |
+
# # allow_origins=["http://localhost:8080"], # Replace with the URL of your Vue.js app
|
| 19 |
+
# allow_origins=["http://localhost:8080"], # Replace with the URL of your Vue.js app
|
| 20 |
+
# allow_credentials=True,
|
| 21 |
+
# allow_methods=["*"], # Allows all HTTP methods (GET, POST, etc.)
|
| 22 |
+
# allow_headers=["*"], # Allows all headers (such as Content-Type, Authorization, etc.)
|
| 23 |
+
# )
|
| 24 |
+
|
| 25 |
+
# # Charger le processor et le modèle fine-tuné depuis le chemin local
|
| 26 |
+
# local_model_path = r'./vit-finetuned-ucf101'
|
| 27 |
+
# processor = AutoImageProcessor.from_pretrained("google/vit-base-patch16-224")
|
| 28 |
+
# model = AutoModelForImageClassification.from_pretrained(local_model_path)
|
| 29 |
+
# # model = AutoModelForImageClassification.from_pretrained("2nzi/vit-finetuned-ucf101")
|
| 30 |
+
# model.eval()
|
| 31 |
+
|
| 32 |
+
# # Fonction pour classifier une image
|
| 33 |
+
# def classifier_image(image):
|
| 34 |
+
# image_pil = Image.fromarray(cv2.cvtColor(image, cv2.COLOR_BGR2RGB))
|
| 35 |
+
# inputs = processor(images=image_pil, return_tensors="pt")
|
| 36 |
+
# with torch.no_grad():
|
| 37 |
+
# outputs = model(**inputs)
|
| 38 |
+
# logits = outputs.logits
|
| 39 |
+
# predicted_class_idx = logits.argmax(-1).item()
|
| 40 |
+
# predicted_class = model.config.id2label[predicted_class_idx]
|
| 41 |
+
# return predicted_class
|
| 42 |
+
|
| 43 |
+
# # Fonction pour traiter la vidéo et identifier les séquences de "Surfing"
|
| 44 |
+
# def identifier_sequences_surfing(video_path, intervalle=0.5):
|
| 45 |
+
# cap = cv2.VideoCapture(video_path)
|
| 46 |
+
# frame_rate = cap.get(cv2.CAP_PROP_FPS)
|
| 47 |
+
# total_frames = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))
|
| 48 |
+
# frame_interval = int(frame_rate * intervalle)
|
| 49 |
+
|
| 50 |
+
# resultats = []
|
| 51 |
+
# sequences_surfing = []
|
| 52 |
+
# frame_index = 0
|
| 53 |
+
# in_surf_sequence = False
|
| 54 |
+
# start_timestamp = None
|
| 55 |
+
|
| 56 |
+
# with tqdm(total=total_frames, desc="Traitement des frames de la vidéo", unit="frame") as pbar:
|
| 57 |
+
# success, frame = cap.read()
|
| 58 |
+
# while success:
|
| 59 |
+
# if frame_index % frame_interval == 0:
|
| 60 |
+
# timestamp = round(frame_index / frame_rate, 2) # Maintain precision to the centisecond level
|
| 61 |
+
# classe = classifier_image(frame)
|
| 62 |
+
# resultats.append({"Timestamp": timestamp, "Classe": classe})
|
| 63 |
+
|
| 64 |
+
# if classe == "Surfing" and not in_surf_sequence:
|
| 65 |
+
# in_surf_sequence = True
|
| 66 |
+
# start_timestamp = timestamp
|
| 67 |
+
|
| 68 |
+
# elif classe != "Surfing" and in_surf_sequence:
|
| 69 |
+
# # Vérifier l'image suivante pour confirmer si c'était une erreur ponctuelle
|
| 70 |
+
# success_next, frame_next = cap.read()
|
| 71 |
+
# next_timestamp = round((frame_index + frame_interval) / frame_rate, 2)
|
| 72 |
+
# classe_next = None
|
| 73 |
+
|
| 74 |
+
# if success_next:
|
| 75 |
+
# classe_next = classifier_image(frame_next)
|
| 76 |
+
# resultats.append({"Timestamp": next_timestamp, "Classe": classe_next})
|
| 77 |
+
|
| 78 |
+
# # Si l'image suivante est "Surfing", on ignore l'erreur ponctuelle
|
| 79 |
+
# if classe_next == "Surfing":
|
| 80 |
+
# success = success_next
|
| 81 |
+
# frame = frame_next
|
| 82 |
+
# frame_index += frame_interval
|
| 83 |
+
# pbar.update(frame_interval)
|
| 84 |
+
# continue
|
| 85 |
+
# else:
|
| 86 |
+
# # Sinon, terminer la séquence "Surfing"
|
| 87 |
+
# in_surf_sequence = False
|
| 88 |
+
# end_timestamp = timestamp
|
| 89 |
+
# sequences_surfing.append((start_timestamp, end_timestamp))
|
| 90 |
+
|
| 91 |
+
# success, frame = cap.read()
|
| 92 |
+
# frame_index += 1
|
| 93 |
+
# pbar.update(1)
|
| 94 |
+
|
| 95 |
+
# # Si on est toujours dans une séquence "Surfing" à la fin de la vidéo
|
| 96 |
+
# if in_surf_sequence:
|
| 97 |
+
# sequences_surfing.append((start_timestamp, round(frame_index / frame_rate, 2)))
|
| 98 |
+
|
| 99 |
+
# cap.release()
|
| 100 |
+
# dataframe_sequences = pd.DataFrame(sequences_surfing, columns=["Début", "Fin"])
|
| 101 |
+
# return dataframe_sequences
|
| 102 |
+
|
| 103 |
+
# # Fonction pour convertir les séquences en format JSON
|
| 104 |
+
# def convertir_sequences_en_json(dataframe):
|
| 105 |
+
# events = []
|
| 106 |
+
# blocks = []
|
| 107 |
+
# for idx, row in dataframe.iterrows():
|
| 108 |
+
# block = {
|
| 109 |
+
# "id": f"Surfing{idx + 1}",
|
| 110 |
+
# "start": round(row["Début"], 2),
|
| 111 |
+
# "end": round(row["Fin"], 2)
|
| 112 |
+
# }
|
| 113 |
+
# blocks.append(block)
|
| 114 |
+
# event = {
|
| 115 |
+
# "event": "Surfing",
|
| 116 |
+
# "blocks": blocks
|
| 117 |
+
# }
|
| 118 |
+
# events.append(event)
|
| 119 |
+
# return events
|
| 120 |
+
|
| 121 |
+
# @app.post("/analyze_video/")
|
| 122 |
+
# async def analyze_video(file: UploadFile = File(...)):
|
| 123 |
+
# with open("uploaded_video.mp4", "wb") as buffer:
|
| 124 |
+
# shutil.copyfileobj(file.file, buffer)
|
| 125 |
+
|
| 126 |
+
# dataframe_sequences = identifier_sequences_surfing("uploaded_video.mp4", intervalle=1)
|
| 127 |
+
# json_result = convertir_sequences_en_json(dataframe_sequences)
|
| 128 |
+
# return json_result
|
| 129 |
+
|
| 130 |
+
# @app.get("/", response_class=HTMLResponse)
|
| 131 |
+
# async def index():
|
| 132 |
+
# return (
|
| 133 |
+
# """
|
| 134 |
+
# <html>
|
| 135 |
+
# <body>
|
| 136 |
+
# <h1>Hello world!</h1>
|
| 137 |
+
# <p>This `/` is the most simple and default endpoint.</p>
|
| 138 |
+
# <p>If you want to learn more, check out the documentation of the API at
|
| 139 |
+
# <a href='/docs'>/docs</a> or
|
| 140 |
+
# <a href='https://2nzi-video-sequence-labeling.hf.space/docs' target='_blank'>external docs</a>.
|
| 141 |
+
# </p>
|
| 142 |
+
# </body>
|
| 143 |
+
# </html>
|
| 144 |
+
# """
|
| 145 |
+
# )
|
| 146 |
+
|
| 147 |
+
|
| 148 |
+
# # Lancer l'application avec uvicorn (command line)
|
| 149 |
+
# # uvicorn main:app --reload
|
| 150 |
+
# # http://localhost:8000/docs#/
|
| 151 |
+
# # (.venv) PS C:\Users\antoi\Documents\Work_Learn\Labeling-Deploy\FastAPI> uvicorn main:app --host 0.0.0.0 --port 8000 --workers 1
|
| 152 |
+
|
| 153 |
+
|
| 154 |
from fastapi import FastAPI, UploadFile, File
|
| 155 |
import cv2
|
| 156 |
import torch
|
|
|
|
| 271 |
events.append(event)
|
| 272 |
return events
|
| 273 |
|
| 274 |
+
|
| 275 |
+
import os
|
| 276 |
+
import tempfile
|
| 277 |
+
|
| 278 |
@app.post("/analyze_video/")
|
| 279 |
async def analyze_video(file: UploadFile = File(...)):
|
| 280 |
+
# Utiliser tempfile pour créer un fichier temporaire
|
| 281 |
+
with tempfile.NamedTemporaryFile(delete=False, suffix=".mp4", dir="/tmp") as tmp:
|
| 282 |
+
shutil.copyfileobj(file.file, tmp)
|
| 283 |
+
tmp_path = tmp.name
|
| 284 |
|
| 285 |
+
# Analyser la vidéo
|
| 286 |
+
dataframe_sequences = identifier_sequences_surfing(tmp_path, intervalle=1)
|
| 287 |
json_result = convertir_sequences_en_json(dataframe_sequences)
|
| 288 |
+
|
| 289 |
+
# Supprimer le fichier temporaire après utilisation
|
| 290 |
+
os.remove(tmp_path)
|
| 291 |
+
|
| 292 |
+
return {"filename": file.filename, "result": json_result}
|
| 293 |
|
| 294 |
@app.get("/", response_class=HTMLResponse)
|
| 295 |
async def index():
|