SM+: Refined Scale Match for Tiny Person Detection
This work provides a strong specific gain for computer vision researchers and practitioners working on tiny object detection, an incremental improvement to existing methods.
This paper addresses the challenge of detecting tiny objects (less than 20x20 pixels) by proposing SM+, a refined Scale Match method that aligns scales between pre-training and target datasets at the instance level. It significantly improves performance on the TinyPerson dataset, outperforming state-of-the-art detectors.
Detecting tiny objects ( e.g., less than 20 x 20 pixels) in large-scale images is an important yet open problem. Modern CNN-based detectors are challenged by the scale mismatch between the dataset for network pre-training and the target dataset for detector training. In this paper, we investigate the scale alignment between pre-training and target datasets, and propose a new refined Scale Match method (termed SM+) for tiny person detection. SM+ improves the scale match from image level to instance level, and effectively promotes the similarity between pre-training and target dataset. Moreover, considering SM+ possibly destroys the image structure, a new probabilistic structure inpainting (PSI) method is proposed for the background processing. Experiments conducted across various detectors show that SM+ noticeably improves the performance on TinyPerson, and outperforms the state-of-the-art detectors with a significant margin.