Object-Pose Estimation With Neural Population Codes

TL;DR

This paper introduces a neural population code approach for object-pose estimation that handles symmetry ambiguities, enabling fast and accurate inference suitable for robotic assembly tasks.

Contribution

It presents a novel neural population coding method that directly maps sensory input to object rotation, overcoming symmetry challenges and reducing computational overhead.

Findings

Achieves 3.2 ms inference time on Apple M1 CPU

Attains 84.7% accuracy on T-LESS dataset

Outperforms direct pose mapping in accuracy

Abstract

Robotic assembly tasks require object-pose estimation, particularly for tasks that avoid costly mechanical constraints. Object symmetry complicates the direct mapping of sensory input to object rotation, as the rotation becomes ambiguous and lacks a unique training target. Some proposed solutions involve evaluating multiple pose hypotheses against the input or predicting a probability distribution, but these approaches suffer from significant computational overhead. Here, we show that representing object rotation with a neural population code overcomes these limitations, enabling a direct mapping to rotation and end-to-end learning. As a result, population codes facilitate fast and accurate pose estimation. On the T-LESS dataset, we achieve inference in 3.2 milliseconds on an Apple M1 CPU and a Maximum Symmetry-Aware Surface Distance accuracy of 84.7% using only gray-scale image input, compared to 69.7% accuracy when directly mapping to pose.