Abstract:Compared with two-stage object detection algorithms, one-stage algorithms provide a better trade-off between real-time performance and accuracy. However, these methods treat the intermediate features equally, which lacks the flexibility to emphasize meaningful information for classification and location. Besides, they ignore the interaction of contextual information from different scales, which is important for medium and small objects detection. To tackle these problems, we propose an image pyramid network based on dual attention mechanism (DAIPNet), which builds an image pyramid to enrich the spatial information while emphasizing multi-scale informative features based on dual attention mechanisms for one stage object detection. Our framework utilizes a pre-trained backbone as standard detection network, where the designed image pyramid network (IPN) is used as auxiliary network to provide complementary information. Here, the dual attention mechanism is composed of the adaptive feature fusion module (AFFM) and the progressive attention fusion module (PAFM). AFFM is designed to automatically pay attention to the feature maps with different importance from the backbone and auxiliary network, while PAFM is utilized to adaptively learn the channel attentive information in the context transfer process. Furthermore, in the IPN, we build an image pyramid to extract scale-wise features from downsampled images of different scales, where the features are further fused at different states to enrich scale-wise information and learn more comprehensive feature representations. Experimental results are shown on MS COCO dataset. Our proposed detector with a 300 x 300 input achieves superior performance of 32.6% mAP on the MS COCO test-dev compared with state-of-the-art methods.

Corner Proposal Network for Anchor-Free, Two-Stage Object Detection

An anchor-free object detector with novel corner matching method

CornerNet: Detecting Objects as Paired Keypoints

Corner2Net: Detecting Objects as Cascade Corners

CenterNet: Keypoint Triplets for Object Detection

CenterNet3D: An Anchor Free Object Detector for Point Cloud

Learning from Noisy Anchors for One-Stage Object Detection.

CPN: Complementary Proposal Network for Unconstrained Text Detection

Multi-modal 3D object detection by 2D-guided precision anchor proposal and multi-layer fusion

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

FCOS: A simple and strong anchor-free object detector

RFP-Net: Receptive field-based proposal generation network for object detection

Robust object proposals re-ranking for object detection in autonomous driving using convolutional neural networks

CenterPoint-SE: A Single-Stage Anchor-Free 3-D Object Detection Algorithm With Spatial Awareness Enhancement

DIGCN: A Dynamic Interaction Graph Convolutional Network Based on Learnable Proposals for Object Detection

Object detection based on scene understanding and enhanced proposals

Toward Scale-Invariance and Position-Sensitive Region Proposal Networks

Dual Attention Based Image Pyramid Network for Object Detection.

Gradient Corner Pooling for Keypoint-Based Object Detection

Anchor-Free Oriented Proposal Generator for Object Detection