Add video person detection tools and comprehensive tutorials #239
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
- Add detect_person_in_video.py: Full-featured video person detection script - Support auto-download from Hugging Face - Support person-only filtering - Support real-time visualization - Support batch processing - Add quick_start_person_detection.py: Simplified quick-start script - Easy configuration in code - Minimal parameters needed - Step-by-step progress display - Add PERSON_DETECTION_TUTORIAL.md: Comprehensive English tutorial - Environment setup guide - Model download instructions (3 methods) - Usage examples (basic to advanced) - Python API documentation - Performance optimization tips - FAQ section - Add QUICK_START_CN.md: Chinese quick-start guide - 3-step quick start - Common usage examples - Parameter reference - Code examples - Troubleshooting Features: - Person detection from video using Co-DETR ViT-Large - Confidence threshold filtering - GPU/CPU support - Real-time preview - Video output with bounding boxes - Detection statistics
Fix: Use correct model filename from HuggingFace - Changed checkpoint filename from 'co_dino_5scale_vit_large_coco.pth' to 'pytorch_model.pth' - Use local config file: projects/configs/co_dino_vit/co_dino_5scale_vit_large_coco.py - Updated all scripts and documentation Updated files: - detect_person_in_video.py: Update download function to use pytorch_model.pth - quick_start_person_detection.py: Update file paths in configuration - PERSON_DETECTION_TUTORIAL.md: Update all download commands and examples - QUICK_START_CN.md: Update Chinese quick-start guide Added: - DOWNLOAD_MODEL.md: Comprehensive model download guide - 3 download methods (CLI, git clone, manual) - File structure explanation - Usage examples - FAQ section Note: The HuggingFace model repository uses 'pytorch_model.pth' as the checkpoint filename, not 'co_dino_5scale_vit_large_coco.pth'
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Add detect_person_in_video.py: Full-featured video person detection script
Add quick_start_person_detection.py: Simplified quick-start script
Add PERSON_DETECTION_TUTORIAL.md: Comprehensive English tutorial
Add QUICK_START_CN.md: Chinese quick-start guide
Features: