Update breathing model to new checkpoint + add locked event-extraction config (threshold=0.03, min_event=0.75s, max_event=4.0s, min_inter_event=1.0s, split-aware)

Update breathing model to v3 (native-fps split, new active-learning round)

Update breathing model to v3 (native-fps split, new active-learning round)

Update breathing model to v3 (native-fps split, new active-learning round)

Update breathing model to v3 (native-fps split, new active-learning round)

Deploy retrained breathing model (temporal-split-corrected dataset v2); optimal_threshold.json now uses the balanced prevalence-adjusted operating point (0.684) instead of leaky-test-set F1-optimal (0.133)

Document new yolo11m-seg checkpoint and inference_config.json

Add calibrated inference_config.json alongside best.pt (conf=0.40, iou=0.45, max_det=1)

Fix: replace flawed yolo11m-seg@640 checkpoint (89 FP/80% recall at real inference settings) with a clean retrain under corrected checkpoint-selection code and corrected val labels. See inference_config.json for the calibrated conf/iou/max_det this checkpoint needs.

Update detector: yolo11m-seg @ imgsz=640, fitness 0.9147 (best result of full resolution+architecture sweep, 2026-08-29/30)

Upload README.md with huggingface_hub

Upload models/breathing/optimal_threshold.json with huggingface_hub

Upload models/breathing/best_model.keras with huggingface_hub

Upload models/yolo/best.pt with huggingface_hub

Upload 1min_sample/D03/9_8_2025/D03_20250809060000.mp4 with huggingface_hub

Upload 1min_sample/D03/8_8_2025/D03_20250808060000.mp4 with huggingface_hub

Upload 1min_sample/D01/9_8_2025/D01_20250809060000.mp4 with huggingface_hub

Upload 1min_sample/D01/8_8_2025/D01_20250808060000.mp4 with huggingface_hub

initial commit