Self-Supervised Learning for Transparent Object Depth Completion Using Depth from Non-Transparent Objects

TL;DR

This paper introduces a self-supervised approach for depth completion of transparent objects, reducing reliance on costly labeled data and achieving performance comparable to supervised methods.

Contribution

The authors propose a novel self-supervised training method that simulates transparent object depth deficits using non-transparent regions, enhancing depth completion without extensive annotations.

Findings

Achieves comparable performance to supervised methods.

Pre-training with our method improves results with limited data.

Effectively simulates transparent object depth deficits.

Abstract

The perception of transparent objects is one of the well-known challenges in computer vision. Conventional depth sensors have difficulty in sensing the depth of transparent objects due to refraction and reflection of light. Previous research has typically train a neural network to complete the depth acquired by the sensor, and this method can quickly and accurately acquire accurate depth maps of transparent objects. However, previous training relies on a large amount of annotation data for supervision, and the labeling of depth maps is costly. To tackle this challenge, we propose a new self-supervised method for training depth completion networks. Our method simulates the depth deficits of transparent objects within non-transparent regions and utilizes the original depth map as ground truth for supervision. Experiments demonstrate that our method achieves performance comparable to supervised approach, and pre-training with our method can improve the model performance when the training samples are small.