Remote sensing images pose significant challenges for object detection due to their high resolution, complex backgrounds, large scale variations, and dense distribution of small targets. To address ...

We introduce YOLO-IOD, an integrated and real-time IOD framework, and pinpoint three causes of forgetting: foreground-background confusion, parameter interference, and misaligned knowledge ...

Start with a foggy image. Apply a dehazing model to remove fog. Pass the dehazed image through YOLO for object detection. Compare the original foggy image, the dehazed result, and the YOLO-detected ...

Abstract: Small object detection faces multiple challenges, including high-frequency information degradation, complex back ground interference, and low target resolution, which frequently lead to ...