Deformable-Heatmap-Segmentation for Automobile Visual Perception

TL;DR

This paper introduces DHSNet, an end-to-end deformable heatmap segmentation network for accurate semantic segmentation of road elements in 2D images, enhancing detection of static objects with varying shapes and scales.

Contribution

The paper presents DHSNet, a novel architecture combining deformable convolutions and heatmap proposals for improved road element segmentation in autonomous driving.

Findings

Effective segmentation of road elements achieved

Improved detection accuracy for objects of various shapes and scales

Enhanced target localization with heatmap proposals

Abstract

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.