Facing depth estimation in-the-wild with deep networks

Talks & Slides

Introduction

Introducing the content of this tutorial, highlighting motivations and challenges[TALK] [SLIDES]

Self-supervised stereo and adaptation

How to train stereo networks without ground truth and how to adapt them to ever-changing environments[TALK] [SLIDES]

Robust stereo and generalization

Effective design strategies to avoid over-specialization in favor of cross-domain generalization[TALK] [SLIDES]

Confidence estimation, latest trends

How confidence estimation can improve a stereo algorithm/network and how to deploy it in-the-wild[TALK] [SLIDES]