TransNet: Category-Level Transparent Object Pose Estimation

TL;DR

TransNet is a novel two-stage method for category-level pose estimation of transparent objects, overcoming visual and depth sensing challenges by using localized depth completion and surface normal estimation, achieving improved accuracy.

Contribution

The paper introduces TransNet, the first approach specifically designed for category-level transparent object pose estimation, addressing unique perception challenges.

Findings

TransNet outperforms existing methods on a large-scale transparent object dataset.

Localized depth completion improves pose estimation accuracy.

Ablation studies highlight key factors influencing performance.

Abstract

Transparent objects present multiple distinct challenges to visual perception systems. First, their lack of distinguishing visual features makes transparent objects harder to detect and localize than opaque objects. Even humans find certain transparent surfaces with little specular reflection or refraction, e.g. glass doors, difficult to perceive. A second challenge is that common depth sensors typically used for opaque object perception cannot obtain accurate depth measurements on transparent objects due to their unique reflective properties. Stemming from these challenges, we observe that transparent object instances within the same category (e.g. cups) look more similar to each other than to ordinary opaque objects of that same category. Given this observation, the present paper sets out to explore the possibility of category-level transparent object pose estimation rather than instance-level pose estimation. We propose TransNet, a two-stage pipeline that learns to estimate category-level transparent object pose using localized depth completion and surface normal estimation. TransNet is evaluated in terms of pose estimation accuracy on a recent, large-scale transparent object dataset and compared to a state-of-the-art category-level pose estimation approach. Results from this comparison demonstrate that TransNet achieves improved pose estimation accuracy on transparent objects and key findings from the included ablation studies suggest future directions for performance improvements.