Skip to yearly menu bar Skip to main content


Poster

Robust Synthetic-to-Real Transfer for Stereo Matching

Jiawei Zhang · Jiahe Li · Lei Huang · Xiaohan Yu · Lin Gu · Jin Zheng · Xiao Bai

Arch 4A-E Poster #69
[ ] [ Paper PDF ]
[ Poster
Fri 21 Jun 10:30 a.m. PDT — noon PDT

Abstract:

With advancements in domain generalized stereo matching networks, models pre-trained on synthetic data demonstrate strong robustness to unseen domains. However, few studies have investigated the robustness after fine-tuning them in real-world scenarios, during which the domain generalization ability can be seriously degraded. In this paper, we explore fine-tuning stereo matching networks without compromising their robustness to unseen domains. Our motivation stems from comparing Ground Truth (GT) versus Pseudo Label (PL) for fine-tuning: GT degrades, but PL preserves the domain generalization ability. Empirically, we find the difference between GT and PL implies valuable information that can regularize networks during fine-tuning. We also propose a framework to utilize this difference for fine-tuning, consisting of a frozen Teacher, an exponential moving average (EMA) Teacher, and a Student network. The core idea is to utilize the EMA Teacher to measure what the Student has learned and dynamically improve GT and PL for fine-tuning. We integrate our framework with state-of-the-art networks and evaluate its effectiveness on several real-world datasets. Extensive experiments show that our method effectively preserves the domain generalization ability during fine-tuning.

Chat is not available.