MFF: An effective method of solving the ill regions in stereo matching

In the current stereo matching field, the accuracy of the derived disparity map is highly dependent on the processing capability of ill regions. Fortunately, we find that the use of local information will eliminate the negative effects associated with ill regions. As a result of the above discovery, we propose the Concatenated Dilated Convolution (CDC) block and the Multi-scale Feature Fusion module (MFF), which are capable of effectively extracting regional context by increasing the receptive field in parallel and channel-wise ways. The CDC block can expand the receptive field by applying multiple dilated convolutions at different dilation rates in parallel to enhance the smoothness of the feature map. By constructing parallel CDC blocks in a multiple dilated manner, the MFF module can improve the smoothness of the feature map. In addition, to control the number of parameters in the MFF network, a high-performance channel-distribution algorithm is proposed, capable of adjusting the weights of each module and convolution in an adaptive manner while reducing the number of parameters. Extensive experiments have demonstrated that MFF and CDC can effectively improve the performance of ill areas and networks with a minimal number of parameters.

Automated summary

In the field of stereo matching, the accuracy of the disparity map depends on how well the algorithm handles ill regions. The proposed Concatenated Dilated Convolution (CDC) block and Multi-scale Feature Fusion module (MFF) effectively extract regional context by increasing the receptive field and enhancing the smoothness of the feature map. Additionally, a channel-distribution algorithm is used to reduce the number of parameters while maintaining performance. Experimental results demonstrate that MFF and CDC improve the performance of ill areas and networks with minimal parameters.