V3Test / image_captioning.py
assentian1970's picture
Update image_captioning.py
d0eb201 verified
import torch
from transformers import AutoModel, AutoTokenizer
from modelscope.hub.snapshot_download import snapshot_download
from PIL import Image
from functools import lru_cache
from decord import VideoReader, cpu
import os
import gc
import cv2
import tempfile
import shutil
import subprocess
import ffmpeg # Added for ffmpeg-python
from yolo_detection import is_image, is_video
# Constants for video processing
MAX_NUM_FRAMES = 32 # Reduced from 64 to potentially avoid OOM
# Check if CUDA is available
DEVICE = "cuda" if torch.cuda.is_available() else "cpu"
global TOTAL_CHUNKS
TOTAL_CHUNKS = 1
# Initialize GPU if available
if DEVICE == "cuda":
def debug():
torch.randn(10).cuda()
debug()
# Model configuration
MODEL_NAME = 'iic/mPLUG-Owl3-7B-240728'
MODEL_CACHE_DIR = "/data/models"
os.makedirs(MODEL_CACHE_DIR, exist_ok=True)
# Download and cache the model
try:
model_path = snapshot_download(MODEL_NAME, cache_dir=MODEL_CACHE_DIR)
except Exception as e:
print(f"Error downloading model: {str(e)}")
model_path = os.path.join(MODEL_CACHE_DIR, MODEL_NAME)
# Model configuration and existing functions remain unchanged...
@lru_cache(maxsize=1)
def load_model_and_tokenizer():
"""Load a cached instance of the model and tokenizer"""
print("Loading/Retrieving mPLUG model from cache...")
try:
# Clear GPU memory if using CUDA
if DEVICE == "cuda":
torch.cuda.empty_cache()
gc.collect()
model = AutoModel.from_pretrained(
model_path,
attn_implementation='sdpa',
trust_remote_code=True,
torch_dtype=torch.half,
device_map='auto'
)
tokenizer = AutoTokenizer.from_pretrained(
model_path,
trust_remote_code=True
)
model.eval()
processor = model.init_processor(tokenizer)
return model, tokenizer, processor
except Exception as e:
print(f"Error loading model: {str(e)}")
raise
def process_image(image_path, model, tokenizer, processor, prompt):
"""Process single image with mPLUG model"""
try:
image = Image.open(image_path)
messages = [{
"role": "user",
"content": prompt,
"images": [image]
}]
model_messages = []
images = []
for msg in messages:
content_str = msg["content"]
if "images" in msg and msg["images"]:
content_str += "<|image|>"
images.extend(msg["images"])
model_messages.append({
"role": msg["role"],
"content": content_str
})
model_messages.append({
"role": "assistant",
"content": ""
})
inputs = processor(
model_messages,
images=images,
videos=None
)
inputs.to('cuda')
inputs.update({
'tokenizer': tokenizer,
'max_new_tokens': 100,
'decode_text': True,
})
response = model.generate(**inputs)
return response[0]
except Exception as e:
print(f"Error processing image: {str(e)}")
return "Error processing image"
def process_video_chunk(video_frames, model, tokenizer, processor, prompt):
"""Process a chunk of video frames with mPLUG model"""
messages = [
{
"role": "user",
"content": prompt,
"video_frames": video_frames
}
]
model_messages = []
videos = []
for msg in messages:
content_str = msg["content"]
if "video_frames" in msg and msg["video_frames"]:
content_str += "<|video|>"
videos.append(msg["video_frames"])
model_messages.append({
"role": msg["role"],
"content": content_str
})
model_messages.append({
"role": "assistant",
"content": ""
})
inputs = processor(
model_messages,
images=None,
videos=videos if videos else None
)
inputs.to('cuda')
inputs.update({
'tokenizer': tokenizer,
'max_new_tokens': 100,
'decode_text': True,
})
response = model.generate(**inputs)
del inputs
return response[0]
def split_original_video(video_path, chunk_info):
"""Split original video into chunks using multiple methods with fallbacks for cross-platform reliability"""
original_chunks = []
# Clean the ./tmp directory containing chunks/thumbnails
tmp_dir = os.path.join('.', 'tmp')
if os.path.exists(tmp_dir):
try:
shutil.rmtree(tmp_dir)
os.makedirs(tmp_dir, exist_ok=True) # Recreate for next run
print(f"Cleaned up temporary directory: {tmp_dir}")
except OSError as e:
print(f"Error removing temporary directory {tmp_dir}: {e}")
else:
os.makedirs(tmp_dir)
for chunk in chunk_info:
chunk_id = chunk['chunk_id']
start_time = chunk['start_time']
end_time = chunk['end_time']
output_path = os.path.join(tmp_dir, f"original_chunk_{chunk_id}.mp4")
# Try three different methods in order of preference
chunk_created = False
# Method 1: Try ffmpeg-python library
if not chunk_created:
try:
(
ffmpeg
.input(video_path, ss=start_time, to=end_time)
.output(output_path, c='copy', loglevel="quiet") # Added loglevel quiet
.run(capture_stdout=True, capture_stderr=True)
)
# Check if file exists and is not empty after ffmpeg-python call
if os.path.exists(output_path) and os.path.getsize(output_path) > 0:
chunk_created = True
print(f"Successfully created chunk {chunk_id} using ffmpeg-python")
else:
print(f"ffmpeg-python ran but did not create a valid file for chunk {chunk_id}")
# Optionally raise an exception here if needed, or just let it proceed to next method
except ffmpeg.Error as e: # Catch specific ffmpeg errors
print(f"ffmpeg-python error for chunk {chunk_id}: {e.stderr.decode() if e.stderr else str(e)}, trying OpenCV method")
except Exception as e: # Catch other potential errors like file not found
print(f"ffmpeg-python failed with general error for chunk {chunk_id}: {str(e)}, trying OpenCV method")
# Method 2: Try OpenCV for video splitting (re-encoding)
if not chunk_created:
try:
cap = cv2.VideoCapture(video_path)
if not cap.isOpened():
raise IOError(f"Cannot open video file: {video_path}")
fps = cap.get(cv2.CAP_PROP_FPS)
if fps <= 0: # Handle case where fps is invalid
print(f"Warning: Invalid FPS ({fps}) detected for {video_path}. Using default 30.")
fps = 30.0
width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
fourcc = cv2.VideoWriter_fourcc(*'mp4v')
out = cv2.VideoWriter(output_path, fourcc, fps, (width, height))
# Calculate frame positions
start_frame = int(start_time * fps)
end_frame = int(end_time * fps)
# Set position to start frame
cap.set(cv2.CAP_PROP_POS_FRAMES, start_frame)
current_frame = start_frame
while current_frame < end_frame:
ret, frame = cap.read()
if not ret:
print(f"Warning: Could not read frame {current_frame} for chunk {chunk_id}. Reached end of video early?")
break # Stop if we can't read a frame
out.write(frame)
current_frame += 1
cap.release()
out.release()
# Check if file exists and is not empty after OpenCV call
if os.path.exists(output_path) and os.path.getsize(output_path) > 0:
chunk_created = True
print(f"Successfully created chunk {chunk_id} using OpenCV")
else:
print(f"OpenCV method ran but did not create a valid file for chunk {chunk_id}")
except Exception as e:
print(f"OpenCV method failed for chunk {chunk_id}: {str(e)}, trying subprocess method")
# Clean up potentially empty file created by OpenCV on error
if os.path.exists(output_path):
try:
os.remove(output_path)
except OSError:
pass # Ignore cleanup error
# Method 3: Last resort - Try subprocess with better error handling
if not chunk_created:
try:
cmd = [
'ffmpeg',
'-ss', str(start_time),
'-to', str(end_time),
'-i', video_path,
'-c', 'copy', # Attempt copy first
'-loglevel', 'error', # Reduce log noise
output_path
]
process = subprocess.run(cmd, capture_output=True, text=True, check=False) # Don't check=True initially
if process.returncode != 0 or not os.path.exists(output_path) or os.path.getsize(output_path) == 0:
print(f"Subprocess ffmpeg copy failed for chunk {chunk_id}. Stderr: {process.stderr}. Trying re-encoding.")
# If copy fails, try re-encoding as a fallback within subprocess
cmd_reencode = [
'ffmpeg',
'-ss', str(start_time),
'-to', str(end_time),
'-i', video_path,
# '-c:v', 'libx264', # Example re-encode, adjust as needed
# '-crf', '23',
# '-c:a', 'aac',
'-loglevel', 'error',
output_path
]
# Ensure overwrite if previous attempt created an empty file
if os.path.exists(output_path):
cmd_reencode.insert(1, '-y') # Add overwrite flag
process_reencode = subprocess.run(cmd_reencode, capture_output=True, text=True, check=False)
if process_reencode.returncode != 0:
raise Exception(f"Subprocess ffmpeg re-encode also failed. Stderr: {process_reencode.stderr}")
# Final check after subprocess attempts
if os.path.exists(output_path) and os.path.getsize(output_path) > 0:
chunk_created = True
print(f"Successfully created chunk {chunk_id} using subprocess ffmpeg")
else:
raise Exception("Subprocess ffmpeg failed to create a valid file.")
except FileNotFoundError:
print(f"Subprocess failed for chunk {chunk_id}: 'ffmpeg' command not found. Ensure ffmpeg is installed and in PATH.")
except Exception as e:
print(f"Subprocess method failed for chunk {chunk_id}: {str(e)}")
# Clean up potentially empty file
if os.path.exists(output_path):
try:
os.remove(output_path)
except OSError:
pass
# If any method succeeded, add the chunk to our list
if chunk_created and os.path.exists(output_path):
original_chunks.append(output_path)
else:
print(f"Warning: Failed to create chunk {chunk_id} using all methods, skipping.")
return original_chunks
def encode_video_in_chunks(video_path):
"""Extract frames from a video in chunks and save chunks to disk"""
global TOTAL_CHUNKS
vr = VideoReader(video_path, ctx=cpu(0))
original_fps = vr.get_avg_fps()
sample_fps = round(original_fps / 1) # 1 FPS
frame_idx = [i for i in range(0, len(vr), sample_fps)]
fps = vr.get_avg_fps()
# Create tmp directory if it doesn't exist
tmp_dir = os.path.join('.', 'tmp')
os.makedirs(tmp_dir, exist_ok=True)
# Split frame indices into chunks
chunks = [
frame_idx[i:i + MAX_NUM_FRAMES]
for i in range(0, len(frame_idx), MAX_NUM_FRAMES)
]
# Set global TOTAL_CHUNKS before processing
TOTAL_CHUNKS = len(chunks)
print(f"Total chunks: {TOTAL_CHUNKS}")
# Information about saved chunks
chunk_info = []
for chunk_idx, chunk in enumerate(chunks):
# Get frames for this chunk
frames = vr.get_batch(chunk).asnumpy()
frames_pil = [Image.fromarray(v.astype('uint8')) for v in frames]
# Save chunk as a video file
chunk_path = os.path.join(tmp_dir, f"chunk_{chunk_idx}.mp4")
# Calculate start and end times for this chunk
if chunk:
start_frame = chunk[0]
end_frame = chunk[-1]
start_time = start_frame / fps
end_time = end_frame / fps
# Save chunk info for later use
chunk_info.append({
'chunk_id': chunk_idx,
'path': chunk_path,
'start_time': start_time,
'end_time': end_time,
'start_frame': start_frame,
'end_frame': end_frame,
'original_fps': fps # Use actual fps from video
})
# Use OpenCV to create video from frames
height, width, _ = frames[0].shape
fourcc = cv2.VideoWriter_fourcc(*'mp4v')
out = cv2.VideoWriter(chunk_path, fourcc, fps, (width, height))
for frame in frames:
# Convert RGB to BGR (OpenCV format)
frame_bgr = cv2.cvtColor(frame, cv2.COLOR_RGB2BGR)
out.write(frame_bgr)
out.release()
print(f"Saved chunk {chunk_idx} to {chunk_path}")
yield chunk_idx, frames_pil, chunk_info[-1] if chunk_info else None
# Split original video after processing all chunks
original_chunks = split_original_video(video_path, chunk_info)
def analyze_image_activities(image_path):
"""Analyze construction site image and generate activity description"""
from datetime import datetime, timedelta
try:
# Sample structured response - Replace with actual model processing
return [
{
'time': datetime.now().strftime("%I:%M %p"),
'summary': 'Excavation work in progress',
'objects': ['excavator', 'worker', 'dump-truck']
},
{
'time': (datetime.now() - timedelta(minutes=30)).strftime("%I:%M %p"),
'summary': 'Material loading operation',
'objects': ['loader', 'worker', 'gravel']
}
]
except Exception as e:
print(f"Error analyzing image: {str(e)}")
return [] # Return empty list on error
def generate_thumbnails(video_path, num_chunks):
"""Extract thumbnails for each chunk
Args:
video_path: Path to video file
num_chunks: Number of chunks to generate thumbnails for
"""
vr = VideoReader(video_path, ctx=cpu(0))
thumbnails = []
total_frames = len(vr)
# Create/clear tmp directory in current working directory
tmp_dir = os.path.join('.', 'tmp')
# Remove existing directory if it exists
if os.path.exists(tmp_dir):
shutil.rmtree(tmp_dir)
os.makedirs(tmp_dir, exist_ok=True)
# Calculate frame step size based on number of chunks
frame_step = total_frames // num_chunks
for chunk_idx in range(num_chunks):
# Take frame at start of each chunk
frame_idx = chunk_idx * frame_step
if frame_idx < total_frames:
frame = vr[frame_idx].asnumpy()
img = Image.fromarray(frame)
temp_path = os.path.join(tmp_dir, f"thumbnail_{chunk_idx}.jpg")
img.save(temp_path)
thumbnails.append({
"path": temp_path,
"time": frame_idx/vr.get_avg_fps()
})
return thumbnails
def analyze_video_activities(video_path, model, tokenizer, processor):
"""Analyze video using mPLUG model with chunking"""
global TOTAL_CHUNKS
# try:
# Existing chunk processing
all_activities = []
# Calculate total chunks first
vr = VideoReader(video_path, ctx=cpu(0))
sample_fps = round(vr.get_avg_fps() / 1)
frame_idx = [i for i in range(0, len(vr), sample_fps)]
TOTAL_CHUNKS = len([frame_idx[i:i + MAX_NUM_FRAMES]
for i in range(0, len(frame_idx), MAX_NUM_FRAMES)])
# Generate thumbnails with known chunk count
thumbnails = generate_thumbnails(video_path, num_chunks=TOTAL_CHUNKS)
# Now process chunks
chunk_generator = encode_video_in_chunks(video_path)
for chunk_idx, video_frames, chunk_info in chunk_generator:
prompt = "Analyze this construction site video chunk and describe the activities happening. Focus on construction activities, machinery usage, and worker actions. Include any construction equipment or machinery you can identify."
response = process_video_chunk(video_frames, model, tokenizer, processor, prompt)
print(f"Chunk {chunk_idx}: {response}")
# Map responses to thumbnails
time_start = chunk_idx * MAX_NUM_FRAMES
chunk_thumbnails = [t for t in thumbnails
if time_start <= t['time'] < time_start + MAX_NUM_FRAMES]
# Extract time from frame position
for thumbnail in chunk_thumbnails:
# Calculate timestamp in minutes:seconds format
seconds = int(thumbnail['time'])
minutes = seconds // 60
seconds = seconds % 60
timestamp = f"{minutes:02d}:{seconds:02d}"
# Extract objects using basic text parsing from the response
# In a production system, you might want to use more sophisticated NLP
objects = []
lower_response = response.lower()
possible_objects = ["excavator", "bulldozer", "crane", "truck", "loader",
"worker", "concrete", "scaffold", "beam", "pipe",
"rebar", "formwork", "drill", "grader", "roller"]
for obj in possible_objects:
if obj in lower_response:
objects.append(obj)
activity = {
'time': timestamp,
'timestamp_seconds': thumbnail['time'], # Store raw seconds for sorting
'summary': response,
'objects': objects,
'thumbnail': thumbnail["path"],
'chunk_id': chunk_idx,
'chunk_path': chunk_info['path'] if chunk_info else None
}
all_activities.append(activity)
# Sort activities by timestamp
all_activities.sort(key=lambda x: x['timestamp_seconds'])
return all_activities
# except Exception as e:
# print(f"Error analyzing video: {str(e)}")
# return [] # Maintain consistent return type