rt-RISeg: Real-Time Model-Free Robot Interactive Segmentation for Active Instance-Level Object Understanding

TL;DR

This paper introduces rt-RISeg, a real-time, model-free interactive perception framework that enables robots to segment unseen objects during manipulation tasks by analyzing body frame velocities, significantly improving generalization over traditional static models.

Contribution

The paper presents a novel, fully self-contained, real-time interactive perception method that does not rely on learned models, improving unseen object segmentation in robotic manipulation.

Findings

Achieves 27.5% higher segmentation accuracy than state-of-the-art methods.

Uses body frame velocities for object identification without learned models.

Segmentation masks can enhance foundation model performance.

Abstract

Successful execution of dexterous robotic manipulation tasks in new environments, such as grasping, depends on the ability to proficiently segment unseen objects from the background and other objects. Previous works in unseen object instance segmentation (UOIS) train models on large-scale datasets, which often leads to overfitting on static visual features. This dependency results in poor generalization performance when confronted with out-of-distribution scenarios. To address this limitation, we rethink the task of UOIS based on the principle that vision is inherently interactive and occurs over time. We propose a novel real-time interactive perception framework, rt-RISeg, that continuously segments unseen objects by robot interactions and analysis of a designed body frame-invariant feature (BFIF). We demonstrate that the relative rotational and linear velocities of randomly sampled body frames, resulting from selected robot interactions, can be used to identify objects without any learned segmentation model. This fully self-contained segmentation pipeline generates and updates object segmentation masks throughout each robot interaction without the need to wait for an action to finish. We showcase the effectiveness of our proposed interactive perception method by achieving an average object segmentation accuracy rate 27.5% greater than state-of-the-art UOIS methods. Furthermore, although rt-RISeg is a standalone framework, we show that the autonomously generated segmentation masks can be used as prompts to vision foundation models for significantly improved performance.