Object-Pose Estimation With Neural Population Codes
This addresses a key challenge in robotic assembly by improving pose estimation accuracy and speed, though it is incremental as it builds on existing methods for handling symmetry.
The paper tackles the problem of object-pose estimation in robotics, where object symmetry causes ambiguity in rotation mapping, by using neural population codes to enable direct mapping and end-to-end learning, resulting in 84.7% accuracy on the T-LESS dataset with 3.2 ms inference time.
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.