MSDC-Net: Multi-Scale Dense and Contextual Networks for Automated Disparity Map for Stereo Matching

TL;DR

This paper introduces MSDC-Net, a novel deep learning architecture that combines multi-scale fusion and residual 3D convolutions to improve disparity map prediction accuracy in stereo matching tasks, especially in non-occluded regions.

Contribution

The paper proposes a new multi-scale dense and contextual network architecture for stereo disparity estimation, integrating multi-scale fusion and residual 3D convolutions for enhanced performance.

Findings

Outperforms existing methods on Scene Flow and KITTI datasets

Achieves higher accuracy in non-occluded regions

Demonstrates effective multi-scale feature fusion

Abstract

Disparity prediction from stereo images is essential to computer vision applications including autonomous driving, 3D model reconstruction, and object detection. To predict accurate disparity map, we propose a novel deep learning architecture for detectingthe disparity map from a rectified pair of stereo images, called MSDC-Net. Our MSDC-Net contains two modules: multi-scale fusion 2D convolution and multi-scale residual 3D convolution modules. The multi-scale fusion 2D convolution module exploits the potential multi-scale features, which extracts and fuses the different scale features by Dense-Net. The multi-scale residual 3D convolution module learns the different scale geometry context from the cost volume which aggregated by the multi-scale fusion 2D convolution module. Experimental results on Scene Flow and KITTI datasets demonstrate that our MSDC-Net significantly outperforms other approaches in the non-occluded region.