Lightweight Monocular Depth with a Novel Neural Architecture Search Method

TL;DR

This paper introduces LiDNAS, a novel neural architecture search method that efficiently generates lightweight monocular depth estimation models, outperforming existing NAS approaches in both efficiency and accuracy.

Contribution

The paper proposes a new Assisted Tabu Search for NAS and constructs a search space on a pre-defined backbone, enabling efficient search for compact depth estimation models.

Findings

LiDNAS outperforms state-of-the-art NAS methods in efficiency and accuracy.

Optimized models are 7%-500% more compact than existing methods.

Achieves superior results on NYU-Depth-v2, KITTI, and ScanNet datasets.

Abstract

This paper presents a novel neural architecture search method, called LiDNAS, for generating lightweight monocular depth estimation models. Unlike previous neural architecture search (NAS) approaches, where finding optimized networks are computationally highly demanding, the introduced novel Assisted Tabu Search leads to efficient architecture exploration. Moreover, we construct the search space on a pre-defined backbone network to balance layer diversity and search space size. The LiDNAS method outperforms the state-of-the-art NAS approach, proposed for disparity and depth estimation, in terms of search efficiency and output model performance. The LiDNAS optimized models achieve results superior to compact depth estimation state-of-the-art on NYU-Depth-v2, KITTI, and ScanNet, while being 7%-500% more compact in size, i.e the number of model parameters.