# 6DoF Object Pose Estimation via Differentiable Proxy Voting Loss

arXiv: https://arxiv.org/abs/2002.03923 · HTML: https://mlsanity.com/paper/2002.03923/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 10 Feb 2020.

## Built on by (most cited first, 7 of 7)

- [Deep Learning-Based Object Pose Estimation: A Comprehensive Survey](https://mlsanity.com/paper/2405.07801/) (2024)
- [Deep Learning on Monocular Object Pose Detection and Tracking: A Comprehensive Overview](https://mlsanity.com/paper/2105.14291/) (2021)
- [PR-GCN: A Deep Graph Convolutional Network with Point Refinement for 6D Pose Estimation](https://mlsanity.com/paper/2108.09916/) (2021)
- [GaitStrip: Gait Recognition via Effective Strip-based Feature Representations and Multi-Level Framework](https://mlsanity.com/paper/2203.03966/) (2022)
- [Depth-based 6DoF Object Pose Estimation using Swin Transformer](https://mlsanity.com/paper/2303.02133/) (2023)
- [Exploiting Point-Wise Attention in 6D Object Pose Estimation Based on Bidirectional Prediction](https://mlsanity.com/paper/2308.08518/) (2023)
- [Learning Point Cloud Representations with Pose Continuity for Depth-Based Category-Level 6D Object Pose Estimation](https://mlsanity.com/paper/2508.14358/) (2025)

## Abstract (arXiv)

Estimating a 6DOF object pose from a single image is very challenging due to occlusions or textureless appearances. Vector-field based keypoint voting has demonstrated its effectiveness and superiority on tackling those issues. However, direct regression of vector-fields neglects that the distances between pixels and keypoints also affect the deviations of hypotheses dramatically. In other words, small errors in direction vectors may generate severely deviated hypotheses when pixels are far away from a keypoint. In this paper, we aim to reduce such errors by incorporating the distances between pixels and keypoints into our objective. To this end, we develop a simple yet effective differentiable proxy voting loss (DPVL) which mimics the hypothesis selection in the voting procedure. By exploiting our voting loss, we are able to train our network in an end-to-end manner. Experiments on widely used datasets, i.e., LINEMOD and Occlusion LINEMOD, manifest that our DPVL improves pose estimation performance significantly and speeds up the training convergence.