Spatial Self-Distillation for Object Detection with Inaccurate Bounding Boxes

TL;DR

This paper introduces SSD-Det, a novel object detection method that leverages spatial self-distillation to refine inaccurate bounding boxes, improving detection accuracy under low-quality supervision.

Contribution

The paper proposes a spatial self-distillation framework with SPSD and SISD modules to effectively utilize spatial information for refining bounding boxes in weakly supervised object detection.

Findings

Achieves state-of-the-art results on MS-COCO and VOC datasets.

Effectively refines bounding boxes with noisy annotations.

Demonstrates robustness to low-quality supervision.

Abstract

Object detection via inaccurate bounding boxes supervision has boosted a broad interest due to the expensive high-quality annotation data or the occasional inevitability of low annotation quality (\eg tiny objects). The previous works usually utilize multiple instance learning (MIL), which highly depends on category information, to select and refine a low-quality box. Those methods suffer from object drift, group prediction and part domination problems without exploring spatial information. In this paper, we heuristically propose a \textbf{Spatial Self-Distillation based Object Detector (SSD-Det)} to mine spatial information to refine the inaccurate box in a self-distillation fashion. SSD-Det utilizes a Spatial Position Self-Distillation \textbf{(SPSD)} module to exploit spatial information and an interactive structure to combine spatial information and category information, thus constructing a high-quality proposal bag. To further improve the selection procedure, a Spatial Identity Self-Distillation \textbf{(SISD)} module is introduced in SSD-Det to obtain spatial confidence to help select the best proposals. Experiments on MS-COCO and VOC datasets with noisy box annotation verify our method's effectiveness and achieve state-of-the-art performance. The code is available at https://github.com/ucas-vg/PointTinyBenchmark/tree/SSD-Det.