Abstract:The recent advances in compressing high-accuracy convolutional neural networks (CNNs) have witnessed remarkable progress in real-time object detection. To accelerate detection speed, lightweight detectors always have few convolution layers using a single-path backbone. Single-path architecture, however, involves continuous pooling and downsampling operations, always resulting in coarse and inaccurate feature maps that are disadvantageous to locate objects. On the other hand, due to limited network capacity, recent lightweight networks are often weak in representing large-scale visual data. To address these problems, we present a dual-path network, named DPNet, with a lightweight attention scheme for real-time object detection. The dual-path architecture enables us to extract in parallel high-level semantic features and low-level object details. Although DPNet has a nearly duplicated shape with respect to single-path detectors, the computational costs and model size are not significantly increased. To enhance representation capability, a lightweight self-correlation module (LSCM) is designed to capture global interactions, with only a few computational overheads and network parameters. In the neck, LSCM is extended into a lightweight cross correlation module (LCCM), capturing mutual dependencies among neighboring scale features. We have conducted exhaustive experiments on MS COCO, Pascal VOC 2007, and ImageNet datasets. The experimental results demonstrate that DPNet achieves a state-of-the-art trade off between detection accuracy and implementation efficiency. More specifically, DPNet achieves 31.3% AP on MS COCO test-dev, 82.7% mAP on Pascal VOC 2007 test set, and 41.6% mAP on ImageNet validation set, together with nearly 2.5M model size, 1.04 GFLOPs, and 164 and 196 frames/s (FPS) FPS for [Formula: see text] input images of three datasets.

Object Detection Based on DenseNet and RPN

A Lightweight SE-YOLOv3 Network for Multi-Scale Object Detection in Remote Sensing Imagery.

DPNet: Dual-Path Network for Real-Time Object Detection With Lightweight Attention

Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks

Rich Features and Precise Localization with Region Proposal Network for Object Detection.

Deconv R-Cnn For Small Object Detection On Remote Sensing Images

Multiclass objects detection algorithm using DarkNet-53 and DenseNet for intelligent vehicles

Deep Regionlets: Blended Representation and Deep Learning for Generic Object Detection

C-RPNs: Promoting object detection in real world via a cascade structure of Region Proposal Networks.

Research and Application of Object Recognition Method Based on Depth Neural Network

Deep Adaptive Proposal Network for Object Detection in Optical Remote Sensing Images

Multi-scale Lightweight Neural Network for Real-Time Object Detection.

PVANET: Deep but Lightweight Neural Networks for Real-time Object Detection

Towards Real-Time Object Detection on Embedded Systems.

C-FCN: Corners-based Fully Convolutional Network for Visual Object Detection

Object Detection via Aspect Ratio and Context Aware Region-based Convolutional Networks

A New Feature Pyramid Network for Object Detection

Tiny-DSOD: Lightweight Object Detection for Resource-Restricted Usages

Reverse Densely Connected Feature Pyramid Network For Object Detection

An Efficient Model Compression Method for CNN Based Object Detection

3D Object Detector: A Multiscale Region Proposal Network Based on Autonomous Driving.