Dual Refinement Feature Pyramid Networks for Object Detection
This work offers an incremental improvement to object detection performance for researchers and practitioners using FPN-based models.
This paper addresses the inefficiency of pixel-wise summation in Feature Pyramid Networks (FPNs) for object detection. They propose Dual Refinement Feature Pyramid Networks (DRFPN), which improves two-stage detectors by 1.6 to 2.2 AP on COCO detection and 1.5 to 1.9 AP on COCO segmentation, and one-stage detectors by 1.9 AP for RetinaNet and 1.3 AP for FCOS.
FPN is a common component used in object detectors, it supplements multi-scale information by adjacent level features interpolation and summation. However, due to the existence of nonlinear operations and the convolutional layers with different output dimensions, the relationship between different levels is much more complex, the pixel-wise summation is not an efficient approach. In this paper, we first analyze the design defects from pixel level and feature map level. Then, we design a novel parameter-free feature pyramid networks named Dual Refinement Feature Pyramid Networks (DRFPN) for the problems. Specifically, DRFPN consists of two modules: Spatial Refinement Block (SRB) and Channel Refinement Block (CRB). SRB learns the location and content of sampling points based on contextual information between adjacent levels. CRB learns an adaptive channel merging method based on attention mechanism. Our proposed DRFPN can be easily plugged into existing FPN-based models. Without bells and whistles, for two-stage detectors, our model outperforms different FPN-based counterparts by 1.6 to 2.2 AP on the COCO detection benchmark, and 1.5 to 1.9 AP on the COCO segmentation benchmark. For one-stage detectors, DRFPN improves anchor-based RetinaNet by 1.9 AP and anchor-free FCOS by 1.3 AP when using ResNet50 as backbone. Extensive experiments verifies the robustness and generalization ability of DRFPN. The code will be made publicly available.