Dive Deeper Into Box for Object Detection

TL;DR

This paper introduces DDBNet, a box reorganization method that refines bounding box localization in anchor-free object detection, achieving state-of-the-art results by filtering and adjusting boxes for better accuracy.

Contribution

The paper presents a novel box reorganization technique that improves localization accuracy in anchor-free object detection models.

Findings

Achieves state-of-the-art detection performance.

Effectively filters out drifted boxes.

Enhances box alignment for precise localization.

Abstract

Anchor free methods have defined the new frontier in state-of-the-art object detection researches where accurate bounding box estimation is the key to the success of these methods. However, even the bounding box has the highest confidence score, it is still far from perfect at localization. To this end, we propose a box reorganization method(DDBNet), which can dive deeper into the box for more accurate localization. At the first step, drifted boxes are filtered out because the contents in these boxes are inconsistent with target semantics. Next, the selected boxes are broken into boundaries, and the well-aligned boundaries are searched and grouped into a sort of optimal boxes toward tightening instances more precisely. Experimental results show that our method is effective which leads to state-of-the-art performance for object detection.