ViTA-Seg: Vision Transformer for Amodal Segmentation in Robotics

TL;DR

ViTA-Seg is a real-time Vision Transformer framework that improves robotic bin-picking by accurately segmenting occluded objects, leveraging global attention and synthetic data for enhanced amodal perception.

Contribution

The paper introduces ViTA-Seg, a novel Vision Transformer-based approach with dual architectures for amodal segmentation, and a synthetic dataset for industrial scenarios.

Findings

Achieves high accuracy in amodal and occlusion segmentation

Operates efficiently for real-time robotic applications

Outperforms existing methods on benchmark datasets

Abstract

Occlusions in robotic bin picking compromise accurate and reliable grasp planning. We present ViTA-Seg, a class-agnostic Vision Transformer framework for real-time amodal segmentation that leverages global attention to recover complete object masks, including hidden regions. We proposte two architectures: a) Single-Head for amodal mask prediction; b) Dual-Head for amodal and occluded mask prediction. We also introduce ViTA-SimData, a photo-realistic synthetic dataset tailored to industrial bin-picking scenario. Extensive experiments on two amodal benchmarks, COOCA and KINS, demonstrate that ViTA-Seg Dual Head achieves strong amodal and occlusion segmentation accuracy with computational efficiency, enabling robust, real-time robotic manipulation.