Extended Ultralytics' Hailo backend from detection-only to seven tasks across the complete .pt → ONNX → INT8 → HEF export path. #25254 · #25259 · #25276 · #25280 · #25283 · #25348

Fixed RKNN INT8 exports across detection, segmentation, pose, and OBB that produced valid files but no detections. #25524

Made channels-last the automatic default for native PyTorch inference on x86 CPUs, improving median throughput by 18.43% for YOLO26n and 14.09% for YOLO11n. #25983

Fixed dataset fraction sampling that selected images from a single class, raising YOLO11n-cls top-1 accuracy from 9.89% to 87.64% in the validation run. #25968

Restored ByteTrack's low-confidence recovery step, which was unreachable under the default configuration and discarded detections in the 0.1–0.25 confidence range. #25034

Removed per-object GPU stalls from result processing, making Results.plot() 7.55× faster and cutting segmentation-validator synchronizations from 876 to 2. #25230 · #25853

Found train/test contamination affecting 27.9% of KITTI depth training images; validation found no statistically significant accuracy impact. #25650

Reduced the DataLoader handoff from 114 objects to 9 per batch, improving throughput by up to 1.67× and preventing file-descriptor crashes at high worker counts. #1399

Rewrote the evaluation confidence sweep from O(T·N) to O(N log N), making it 7.1–7.8× faster at COCO scale with identical macro_f1 results. #1339

Reduced Hungarian matcher time and peak CUDA memory by up to 77% by padding to the largest target count in each batch instead of the total number of targets. #1297

Corrected the encoder mask-loss computation across all released segmentation models and added the missing tests for SegmentationHead. #1331

Aligned ONNX/TFLite decoding with PyTorch, restoring 30 detections dropped across 26 of 200 COCO validation images. #1320

Preserved callback state after resuming training, preventing diverged runs from overwriting good checkpoints with NaN weights. #1318

Fixed a float-equality bug that degraded steady-frame-rate tracking by as much as 4.86 HOTA and 236 extra ID switches on DanceTrack. #531

Fixed the CLI, demo, and benchmark paths silently assuming 30 fps—a default that was wrong for every SportsMOT and DanceTrack sequence supported by the benchmark. #569

Fixed OC-SORT omitting low-confidence inputs from its return value, restoring one output row per detection. #566

Reduced a Kalman filter noise-calculation hot path by 64.93% while preserving bit-identical output. #572