Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Video

Updated: Dec 15, 2025

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

Self-Supervised Multiscale Adversarial Regression Network for Stereo Disparity Estimation.

Chen Wang, Xiao Bai, Xiang Wang

IEEE Transactions on Cybernetics

|July 11, 2020

Summary

This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

[Hemodynamic effects of synchronous and asynchronous independent lung ventilation with different levels of positive end-expiratory pressure and tidal volumes on unilateral lung injury in dogs].

Zhonghua jie he he hu xi za zhi = Zhonghua jiehe he huxi zazhi = Chinese journal of tuberculosis and respiratory diseases·2010

Same author

[Study on the immuno-effects and influencing factors of Chinese hamster ovary (CHO) cell hepatitis B vaccine among adults, under different dosages].

Zhonghua liu xing bing xue za zhi = Zhonghua liuxingbingxue zazhi·2010

Same author

Assemblies of fluorine containing bent-shaped liquid crystal molecules studied by using scanning tunneling microscopy.

Journal of nanoscience and nanotechnology·2010

Same author

Carbon nanotubes induce secondary structure changes of bovine albumin in aqueous phase.

Journal of nanoscience and nanotechnology·2010

Same author

[Immunogenicity and protective efficacy of pertactin recombinants against Bordetella bronchiseptica challenge].

Wei sheng wu xue bao = Acta microbiologica Sinica·2010

Same author

[Analysis of the electrocardiographic findings in 288 patients with acute pulmonary thromboembolism].

Zhonghua jie he he hu xi za zhi = Zhonghua jiehe he huxi zazhi = Chinese journal of tuberculosis and respiratory diseases·2010

This study introduces a self-supervised deep learning network (SMAR-Net) for stereo matching, eliminating the need for costly ground-truth depth maps. SMAR-Net achieves state-of-the-art results in self-supervised stereo vision, comparable to supervised methods.

Area of Science:

Computer Vision
Deep Learning
Machine Learning

Background:

Supervised deep learning for stereo matching requires extensive, high-quality ground-truth depth data, which is difficult and expensive to acquire.
Existing stereo vision datasets are limited and often unsuitable for deep supervised training.

Purpose of the Study:

To propose a novel self-supervised deep stereo matching approach (SMAR-Net) that removes the dependency on ground-truth depth maps.
To enhance stereo matching accuracy in challenging, ill-posed regions.

Main Methods:

A two-stage network: a disparity regressor using stereo image stacking and a synthetic image generator based on left-right consistency.
Training involves minimizing a hybrid loss function with content loss (warping error) and a novel multiscale adversarial loss to penalize mismatches.

More Related Videos

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

Related Experiment Videos

Last Updated: Dec 15, 2025

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

Main Results:

SMAR-Net demonstrates superior performance compared to existing self-supervised stereo matching methods.
The method achieves results comparable to traditional supervised approaches on benchmark datasets.
Multiscale feature extraction improves adaptability in ill-posed regions.

Conclusions:

SMAR-Net offers an effective self-supervised solution for stereo matching, overcoming data acquisition limitations.
The proposed network advances the field of stereo vision by enabling high-quality depth estimation without ground truth.