๐๐จ๐ฅ๐ฎ๐ญ๐ข๐จ๐งไธจYOLO Object Detection & Tracking
Liu Chang
๐๐ข๐๐๐จ ๐๐ซ๐๐๐ค๐ข๐ง๐ : Road, low-light and aerial footage. Anonymous IDs and real CSV outputs. Media credits included.
Computer Vision Engineer | Detection, Tracking & Data Export
๐๐จ๐ญ๐ข๐ฏ๐๐ญ๐ข๐จ๐ง & ๐๐ก๐๐ฅ๐ฅ๐๐ง๐ ๐: Turn varied images and videos into usable object and motion data. I leveraged YOLO-based and open-vocabulary detectors across 10 representative scene categories, then adapted prompts and outputs to each use case.
๐๐๐ฌ๐ฎ๐ฅ๐ญ๐ฌ & ๐๐๐ฅ๐ข๐ฏ๐๐ซ๐๐๐ฅ๐๐ฌ: Robustly annotated traffic videos with anonymous IDs, domain-specific detection examples, and timestamped CSV/JSON exports.
๐๐ฉ๐จ๐ซ๐ญ๐ฌ & ๐๐๐จ๐ฉ๐ฅ๐: Player, ball and people detection. Faces blurred. Image examples; no team recognition or player tracking.
๐๐ง๐ข๐ฆ๐๐ฅ๐ฌ: Horse and pasture scenes from saved YOLO predictions. Reviewable boxes; no race-speed or livestock-count accuracy claim.
๐๐ซ๐จ๐๐ฎ๐๐ญ๐ฌ & ๐๐ง๐๐ฎ๐ฌ๐ญ๐ซ๐ฒ: Shelf proposals, crate/pallet and object localization. Human review required; no verified SKU counts.
Like this project
Posted Sep 16, 2026
Robustly annotated traffic videos with anonymous IDs, domain-specific detection examples, and timestamped CSV/JSON exports.
Likes
1
Views
1
