Key-Grid: Unsupervised 3D Keypoints Detection using Grid Heatmap Features

TL;DR

This paper introduces Key-Grid, an unsupervised 3D keypoint detection method that effectively handles both rigid and deformable objects by using a novel grid heatmap representation, achieving state-of-the-art results.

Contribution

The paper proposes a new autoencoder framework with a grid heatmap for unsupervised 3D keypoint detection, extending capabilities to deformable objects and ensuring SE(3) invariance.

Findings

State-of-the-art semantic consistency and position accuracy

Robustness to noise and downsampling

Effective for both rigid and deformable objects

Abstract

Detecting 3D keypoints with semantic consistency is widely used in many scenarios such as pose estimation, shape registration and robotics. Currently, most unsupervised 3D keypoint detection methods focus on the rigid-body objects. However, when faced with deformable objects, the keypoints they identify do not preserve semantic consistency well. In this paper, we introduce an innovative unsupervised keypoint detector Key-Grid for both the rigid-body and deformable objects, which is an autoencoder framework. The encoder predicts keypoints and the decoder utilizes the generated keypoints to reconstruct the objects. Unlike previous work, we leverage the identified keypoint in formation to form a 3D grid feature heatmap called grid heatmap, which is used in the decoder section. Grid heatmap is a novel concept that represents the latent variables for grid points sampled uniformly in the 3D cubic space, where these variables are the shortest distance between the grid points and the skeleton connected by keypoint pairs. Meanwhile, we incorporate the information from each layer of the encoder into the decoder section. We conduct an extensive evaluation of Key-Grid on a list of benchmark datasets. Key-Grid achieves the state-of-the-art performance on the semantic consistency and position accuracy of keypoints. Moreover, we demonstrate the robustness of Key-Grid to noise and downsampling. In addition, we achieve SE-(3) invariance of keypoints though generalizing Key-Grid to a SE(3)-invariant backbone.