DenseDisp: Resource-Aware Disparity Map Estimation by Compressing Siamese Neural Architecture

Stereo vision cameras are flexible sensors due to providing heterogeneous information such as color, luminance, disparity map (depth), and shape of the objects. Today, Convolutional Neural Networks (CNNs) present the highest accuracy for the disparity map estimation [1]. However, CNNs require considerable computing capacity to process billions of floating-point operations in a real-time fashion. Besides, commercial stereo cameras produce huge size images (e.g., 10 Megapixels [2]), which impose a new computational cost to the system. The problem will be pronounced if we target resource-limited hardware for the implementation. In this paper, we propose DenseDisp, an automatic framework that designs a Siamese neural architecture for disparity map estimation in a reasonable time. DenseDisp leverages a meta-heuristic multi-objective exploration to discover hardware-friendly architectures by considering accuracy and network FLOPS as the optimization objectives. We explore the design space with four different fitness functions to improve the accuracy-FLOPS trade-off and convergency time of the DenseDisp. According to the experimental results, DenseDisp provides up to 39.1x compression rate while losing around 5% accuracy compared to the state-of-the-art results.