Computer vision / Mar – May 2025
Jersey Number Recognition Boosted
Sports-video recognition with a bag of tricks
Recognizes jersey numbers across player tracklets using pose-guided crops, scene-text recognition, temporal smoothing, and reduced-precision pose inference.
A bag of tricks
The public repository combines denoising, augmentation, legibility filtering, pose-guided crops, scene-text recognition, temporal smoothing, and learned prediction consolidation. The ViTPose attention implementation is optimized for inference.
What I built
- Built an end-to-end computer vision pipeline to extract jersey numbers from sports video, combining person tracking, pose-guided ROI cropping, and OCR/scene-text recognition to produce per-player number predictions.
- Implemented pose-detection pipeline improvements by adding temporal smoothing for keypoints and enabled GPU reduced-precision inference to improve stability and performance.
- Hardened pipeline execution by improving device selection (CPU/GPU), fixing keypoint serialization for JSON outputs, tuning pose-based cropping thresholds, for reduced latency and reproducible runs.
Explore the implementation
Source code, setup instructions, and project documentation are available in the repository.