Single Pixel Reconstruction for One-stage Instance Segmentation
This addresses the efficiency problem for computer vision practitioners by providing a faster alternative to two-stage instance segmentation, though it is incremental as it builds on existing one-stage detectors.
The paper tackles the inefficiency of two-stage instance segmentation methods by proposing SPRNet, a one-stage framework that adds a single pixel reconstruction branch to existing detectors, achieving comparable mask AP to Mask R-CNN with higher inference speed and improved box AP across scales compared to RetinaNet.
Object instance segmentation is one of the most fundamental but challenging tasks in computer vision, and it requires the pixel-level image understanding. Most existing approaches address this problem by adding a mask prediction branch to a two-stage object detector with the Region Proposal Network (RPN). Although producing good segmentation results, the efficiency of these two-stage approaches is far from satisfactory, restricting their applicability in practice. In this paper, we propose a one-stage framework, SPRNet, which performs efficient instance segmentation by introducing a single pixel reconstruction (SPR) branch to off-the-shelf one-stage detectors. The added SPR branch reconstructs the pixel-level mask from every single pixel in the convolution feature map directly. Using the same ResNet-50 backbone, SPRNet achieves comparable mask AP to Mask R-CNN at a higher inference speed, and gains all-round improvements on box AP at every scale comparing with RetinaNet.