Extract human detections and crops from dashcam videos with YOLOv8 and ByteTrack.
npx skills add https://github.com/yousufjoyian/claude-skills --skill human-extractor
Please help me install this Agent Skill. Skill: human-extractor Source: https://github.com/yousufjoyian/claude-skills/tree/main/human-extractor Command: npx skills add https://github.com/yousufjoyian/claude-skills --skill human-extractor
This Skill automates the detection and extraction of human subjects from dashcam video footage, enabling rapid retrieval of evidence-ready crops and searchable metadata.
Run the extraction workflow by providing a JSON config that points to source dashcam directories and an output folder. Example: create config.json from assets/config_template.json and run:
High-intent search queries and answers about installing and using this skill.
Extract humans from dashcam videos using YOLOv8 person detection paired with ByteTrack for stable tracking across frames. This Skill automates detection, crops detected humans, and generates searchable CSV indexes (per-date MANIFEST.csv and global INDEX.csv) for rapid evidence retrieval.
Yes. This Skill GPU-accelerates the entire pipeline with YOLOv8 and ByteTrack, eliminating manual review. It outputs cropped human images and metadata indexes automatically, enabling fast search across your dashcam dataset.
Processing is idempotent and resume-safe. The Skill skips already-processed crops, so you can interrupt and restart without re-processing completed frames or losing progress.
The Skill builds per-date MANIFEST.csv files and a global INDEX.csv that index all extracted humans. Query the index to retrieve crops and metadata across dates, enabling cross-file searches and audit trails.
Yes. The Skill supports optional CLIP-based head-covering classification in extended pipelines, allowing you to filter or label extracted humans by visual attributes beyond raw detection.
Provide source dashcam video directories and an output folder configured in a JSON file (template in assets/config_template.json). A GPU is required for accelerated YOLOv8 detection and ByteTrack processing.
HyperDbgGenerate Go unit test skeletons from interface definitions.
JadeLiu-techDecomposes images into transparent PNG assets with a JSON manifest via CLI or Python API.
curiositechGuide CLIP image-text matching with decision trees and alternative model recommendations.
zhangzhang-111-iGuide object detection and image segmentation with PyTorch and ONNX deployment.
