Deformable-Heatmap-Segmentation for Automobile Visual Perception
This addresses the problem of recognizing static objects for automobile visual perception, but appears incremental as it builds on existing segmentation methods with specific enhancements.
The paper tackles semantic segmentation of road elements like lane lines and free space in 2D images, proposing DHSNet which uses deformable convolutions and a heatmap proposal to achieve more accurate target aiming.
Semantic segmentation of road elements in 2D images is a crucial task in the recognition of some static objects such as lane lines and free space. In this paper, we propose DHSNet,which extracts the objects features with a end-to-end architecture along with a heatmap proposal. Deformable convolutions are also utilized in the proposed network. The DHSNet finely combines low-level feature maps with high-level ones by using upsampling operators as well as downsampling operators in a U-shape manner. Besides, DHSNet also aims to capture static objects of various shapes and scales. We also predict a proposal heatmap to detect the proposal points for more accurate target aiming in the network.