Abstract:There are still two problems in SDD causing some inaccurate results: (1) In the process of feature extraction, with the layer-by-layer acquisition of semantic information, local information is gradually lost, resulting into less representative feature maps; (2) During the Non-Maximum Suppression (NMS) algorithm due to inconsistency in classification and regression tasks, the classification confidence and predicted detection position cannot accurately indicate the position of the prediction boxes. Methods: In order to address these aforementioned issues, we propose a new architecture, a modified version of Single Shot Multibox Detector (SSD), named Precise Single Stage Detector (PSSD). Firstly, we improve the features by adding extra layers to SSD. Secondly, we construct a simple and effective feature enhancement module to expand the receptive field step by step for each layer and enhance its local and semantic information. Finally, we design a more efficient loss function to predict the IOU between the prediction boxes and ground truth boxes, and the threshold IOU guides classification training and attenuates the scores, which are used by the NMS algorithm. Main Results: Benefiting from the above optimization, the proposed model PSSD achieves exciting performance in real-time. Specifically, with the hardware of Titan Xp and the input size of 320 pix, PSSD achieves 33.8 mAP at 45 FPS speed on MS COCO benchmark and 81.28 mAP at 66 FPS speed on Pascal VOC 2007 outperforming state-of-the-art object detection models. Besides, the proposed model performs significantly well with larger input size. Under 512 pix, PSSD can obtain 37.2 mAP with 27 FPS on MS COCO and 82.82 mAP with 40 FPS on Pascal VOC 2007. The experiment results prove that the proposed model has a better trade-off between speed and accuracy.

Mask-guided SSD for small-object detection

Mask SSD: an Effective Single-Stage Approach to Object Instance Segmentation

MT-SSD: Single-Stage 3D Object Detector Based on Magnification Transformation

SAFF-SSD: Self-Attention Combined Feature Fusion-Based SSD for Small Object Detection in Remote Sensing

Detecting Small Objects in Thermal Images Using Single-Shot Detector

New Default Box Strategy of SSD for Small Target Detection

An Improved SSD for small target detection.

Feature-fused SSD: fast detection for small objects

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

SSD: Single Shot MultiBox Detector

Small-objectness Sensitive Detection Based on Shifted Single Shot Detector

Multi-scales Feature Integration Single Shot Multi-Box Detector on Small Object Detection

Attention Based Single Shot Multibox Detector

Precise Single-stage Detector

Pedestrian Detection Based on Improved SSD Object Detection Algorithm

An effective method for small objects detection based on MDFFAM and LKSPP

L-SSD: lightweight SSD target detection based on depth-separable convolution

A single-shot model for traffic-related pedestrian detection

Cascade Mask Generation Framework for Fast Small Object Detection

FSSD: Feature Fusion Single Shot Multibox Detector

Spatial Enhanced-SSD For Multiclass Object Detection in Remote Sensing Images