Using Self-Contradiction to Learn Confidence Measures in Stereo Vision

TL;DR

This paper introduces a scalable method for generating training data for confidence measures in stereo vision by analyzing contradictions across multiple viewpoints, eliminating the need for manual labeling or synthetic data.

Contribution

It presents a novel automated data generation approach using view point contradictions, significantly reducing manual effort and enabling large-scale training for confidence measures.

Findings

Boosted confidence measure performance on KITTI2012

Generated large-scale training data automatically

Reduced reliance on manual or synthetic data

Abstract

Learned confidence measures gain increasing importance for outlier removal and quality improvement in stereo vision. However, acquiring the necessary training data is typically a tedious and time consuming task that involves manual interaction, active sensing devices and/or synthetic scenes. To overcome this problem, we propose a new, flexible, and scalable way for generating training data that only requires a set of stereo images as input. The key idea of our approach is to use different view points for reasoning about contradictions and consistencies between multiple depth maps generated with the same stereo algorithm. This enables us to generate a huge amount of training data in a fully automated manner. Among other experiments, we demonstrate the potential of our approach by boosting the performance of three learned confidence measures on the KITTI2012 dataset by simply training them on a vast amount of automatically generated training data rather than a limited amount of laser ground truth data.