EASNet: Searching Elastic and Accurate Network Architecture for Stereo Matching

07/20/2022
by   Qiang Wang, et al.
0

Recent advanced studies have spent considerable human efforts on optimizing network architectures for stereo matching but hardly achieved both high accuracy and fast inference speed. To ease the workload in network design, neural architecture search (NAS) has been applied with great success to various sparse prediction tasks, such as image classification and object detection. However, existing NAS studies on the dense prediction task, especially stereo matching, still cannot be efficiently and effectively deployed on devices of different computing capabilities. To this end, we propose to train an elastic and accurate network for stereo matching (EASNet) that supports various 3D architectural settings on devices with different computing capabilities. Given the deployment latency constraint on the target device, we can quickly extract a sub-network from the full EASNet without additional training while the accuracy of the sub-network can still be maintained. Extensive experiments show that our EASNet outperforms both state-of-the-art human-designed and NAS-based architectures on Scene Flow and MPI Sintel datasets in terms of model accuracy and inference speed. Particularly, deployed on an inference GPU, EASNet achieves a new SOTA 0.73 EPE on the Scene Flow dataset with 100 ms, which is 4.5× faster than LEAStereo with a better quality model.

READ FULL TEXT
research
10/26/2020

Hierarchical Neural Architecture Search for Deep Stereo Matching

To reduce the human efforts in neural network design, Neural Architectur...
research
10/06/2022

Inference Latency Prediction at the Edge

With the growing workload of inference tasks on mobile devices, state-of...
research
08/03/2021

Elastic Architecture Search for Diverse Tasks with Different Resources

We study a new challenging problem of efficient deployment for diverse t...
research
08/29/2018

Searching Toward Pareto-Optimal Device-Aware Neural Architectures

Recent breakthroughs in Neural Architectural Search (NAS) have achieved ...
research
05/10/2019

Single-Path NAS: Device-Aware Efficient ConvNet Design

Can we automatically design a Convolutional Network (ConvNet) with the h...
research
04/19/2020

Superkernel Neural Architecture Search for Image Denoising

Recent advancements in Neural Architecture Search(NAS) resulted in findi...
research
07/31/2020

Searching Efficient 3D Architectures with Sparse Point-Voxel Convolution

Self-driving cars need to understand 3D scenes efficiently and accuratel...

Please sign up or login with your details

Forgot password? Click here to reset