FastStereoNet: A Fast Neural Architecture Search for Improving the Inference of Disparity Estimation on Resource-Limited Platforms

Convolutional Neural Networks (CNNs) provide the best accuracy for disparity estimation. However, CNNs are computationally expensive, making them unfavorable for resource- limited devices with real-time constraints. Recent advances in Neural Architectures Search (NAS) promise opportunities in automated optimization for disparity estimation [1], [2]. However, the main challenge of the NAS methods is the significant amount of computing time to explore a vast search space (e.g., 1.6×10e29 [3]) and costly training candidates. To reduce the NAS computational demand, many proxy-based NAS methods have been proposed. Despite their success, most of them are designed for comparatively small-scale learning tasks. In this paper, we propose a fast NAS method, called FastStereoNet, to enable resource-aware NAS within an intractably large search space. FastStereoNet automatically searches for hardware-friendly CNN architectures based on Late Acceptance Hill Climbing (LAHC), followed by Simulated Annealing (SA). FastStereoNet also employs a fine-tuning with transferred weights mechanism to improve the convergence of the search process. Collection of these ideas provides competitive results in terms of search time and strikes a balance between accuracy and efficiency. Compared to the state-of-the-art [1], FastStereoNet provides 5.25× reduction in search time and 44.4× reduction in model size. This benefits are attained while yielding a comparable accuracy that enables seamless deployment of disparity estimation on resource-limited devices. Finally, FastStereoNet significantly improves the perception quality of disparity estimation deployed on FPGA and Intel® NCS2 accelerator in a significantly less onerous manner.