MDSSD: Multi-scale Deconvolutional Single Shot Detector for Small Objects

TL;DR

The paper introduces MDSSD, a multi-scale deconvolutional single shot detector designed to improve small object detection by upsampling high-level features and fusing them with low-level features, achieving state-of-the-art results.

Contribution

It proposes a novel multi-scale deconvolutional framework with Fusion Blocks that enhance small object detection by preserving semantic and fine details.

Findings

Achieves 77.6% mAP on TT100K for small objects

Outperforms other detectors on PASCAL VOC2007 and MS COCO

Improves small object detection by 2-5 points on key datasets

Abstract

For most of the object detectors based on multi-scale feature maps, the shallow layers are rich in fine spatial information and thus mainly responsible for small object detection. The performance of small object detection, however, is still less than satisfactory because of the deficiency of semantic information on shallow feature maps. In this paper, we design a Multi-scale Deconvolutional Single Shot Detector (MDSSD), especially for small object detection. In MDSSD, multiple high-level feature maps at different scales are upsampled simultaneously to increase the spatial resolution. Afterwards, we implement the skip connections with low-level feature maps via Fusion Block. The fusion feature maps, named Fusion Module, are of strong feature representational power of small instances. It is noteworthy that these high-level feature maps utilized in Fusion Block preserve both strong semantic information and some fine details of small instances, rather than the top-most layer where the representation of fine details for small objects are potentially wiped out. The proposed framework achieves 77.6% mAP for small object detection on the challenging dataset TT100K with 512 x 512 input, outperforming other detectors with a large margin. Moreover, it can also achieve state-of-the-art results for general object detection on PASCAL VOC2007 test and MS COCO test-dev2015, especially achieving 2 to 5 points improvement on small object categories.