# 2D Pre-Training for 3D Pose Estimation

arXiv: https://arxiv.org/abs/2604.22830 · HTML: https://mlsanity.com/paper/2604.22830/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 24 of its 35 references that are in the map.
- First posted on arXiv 19 Apr 2026.

## Builds on (most cited first, 24 of 24)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Very Deep Convolutional Networks for Large-Scale Image Recognition](https://mlsanity.com/paper/1409.1556/) (2014)
- [The Unreasonable Effectiveness of Deep Features as a Perceptual Metric](https://mlsanity.com/paper/1801.03924/) (2018)
- [Fast R-CNN](https://mlsanity.com/paper/1504.08083/) (2015)
- [SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size](https://mlsanity.com/paper/1602.07360/) (2016)
- [Strategies for Pre-training Graph Neural Networks](https://mlsanity.com/paper/1905.12265/) (2019)
- [VIBE: Video Inference for Human Body Pose and Shape Estimation](https://mlsanity.com/paper/1912.05656/) (2019)
- [Monocular 3D Human Pose Estimation In The Wild Using Improved CNN Supervision](https://mlsanity.com/paper/1611.09813/) (2016)
- [Integral Human Pose Regression](https://mlsanity.com/paper/1711.08229/) (2017)
- [Semantic Graph Convolutional Networks for 3D Human Pose Regression](https://mlsanity.com/paper/1904.03345/) (2019)
- [Single-Shot Multi-Person 3D Pose Estimation From Monocular RGB](https://mlsanity.com/paper/1712.03453/) (2017)
- [Long-term Human Motion Prediction with Scene Context](https://mlsanity.com/paper/2007.03672/) (2020)
- [Compositional Human Pose Regression](https://mlsanity.com/paper/1704.00159/) (2017)
- [Towards 3D Human Pose Estimation in the Wild: a Weakly-supervised Approach](https://mlsanity.com/paper/1704.02447/) (2017)
- [2D/3D Pose Estimation and Action Recognition using Multitask Deep Learning](https://mlsanity.com/paper/1802.09232/) (2018)
- [Self-Supervised Learning of 3D Human Pose using Multi-view Geometry](https://mlsanity.com/paper/1903.02330/) (2019)
- [Monocular 3D Human Pose Estimation by Generation and Ordinal Ranking](https://mlsanity.com/paper/1904.01324/) (2019)
- [Fast Human Pose Estimation](https://mlsanity.com/paper/1811.05419/) (2018)
- [UniPose: Unified Human Pose Estimation in Single Images and Videos](https://mlsanity.com/paper/2001.08095/) (2020)
- [AdaFuse: Adaptive Multiview Fusion for Accurate Human Pose Estimation in the Wild](https://mlsanity.com/paper/2010.13302/) (2020)
- [Reconstructing NBA Players](https://mlsanity.com/paper/2007.13303/) (2020)
- [Combining Local Appearance and Holistic View: Dual-Source Deep Neural Networks for Human Pose Estimation](https://mlsanity.com/paper/1504.07159/) (2015)
- [AIO-P: Expanding Neural Performance Predictors Beyond Image Classification](https://mlsanity.com/paper/2211.17228/) (2022)
- [An Integral Pose Regression System for the ECCV2018 PoseTrack Challenge](https://mlsanity.com/paper/1809.06079/) (2018)

## Abstract (arXiv)

Pre-training is a general method that is used in a range of deep learning tasks. By first training a model on one task, and then further training on the downstream task used for final evaluation, the model is forced to learn a more general understanding of the input data. While pre-training has been applied to 3D Human Pose Estimation (HPE) previously, the scope of datasets used is typically very limited to some strong benchmarks, like Human3.6M. Therefore, in this project, we expand the scope of an existing 3D HPE scheme to be compatible with additional 2D and 3D HPE datasets, like Occlusion Person. We perform an extensive study on how aspects of 2D pre-training, such as model size, affect downstream performance, and to what extent pre-training can help the model generalize to different datasets. Experimental results show that 2D pre-training consistently outperforms training on 3D data alone, particularly in terms of computational efficiency. Finally, using MPII and Human3.6M, we are able to obtain an MPJPE score of under 64.5mm.