# STA-TFM: Spatio-Temporal Aggregation Across Views TransForMer for Pose Estimation

arXiv: https://arxiv.org/abs/2609.24482 · HTML: https://mlsanity.com/paper/2609.24482/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 21 of its 37 references that are in the map.
- First posted on arXiv 21 Sep 2026.

## Builds on (most cited first, 21 of 21)

- [3D human pose estimation in video with temporal convolutions and semi-supervised training](https://mlsanity.com/paper/1811.11742/) (2018)
- [Deep Learning-Based Human Pose Estimation: A Survey](https://mlsanity.com/paper/2012.13392/) (2020)
- [MotionBERT: A Unified Perspective on Learning Human Motion Representations](https://mlsanity.com/paper/2210.06551/) (2022)
- [RTMPose: Real-Time Multi-Person Pose Estimation based on MMPose](https://mlsanity.com/paper/2303.07399/) (2023)
- [Panoptic Studio: A Massively Multiview System for Social Interaction Capture](https://mlsanity.com/paper/1612.03153/) (2016)
- [Epipolar Transformers](https://mlsanity.com/paper/2005.04551/) (2020)
- [BlazePose: On-device Real-time Body Pose tracking](https://mlsanity.com/paper/2006.10204/) (2020)
- [PoseFormerV2: Exploring Frequency Domain for Efficient and Robust 3D Human Pose Estimation](https://mlsanity.com/paper/2303.17472/) (2023)
- [Cascaded deep monocular 3D human pose estimation with evolutionary training data](https://mlsanity.com/paper/2006.07778/) (2020)
- [Direct Multi-view Multi-person 3D Pose Estimation](https://mlsanity.com/paper/2111.04076/) (2021)
- [MotionAGFormer: Enhancing 3D Human Pose Estimation with a Transformer-GCNFormer Network](https://mlsanity.com/paper/2310.16288/) (2023)
- [TransFusion: Cross-view Fusion with Transformer for 3D Human Pose Estimation](https://mlsanity.com/paper/2110.09554/) (2021)
- [TEMPO: Efficient Multi-View Pose Estimation, Tracking, and Forecasting](https://mlsanity.com/paper/2309.07910/) (2023)
- [SelfPose3d: Self-Supervised Multi-Person Multi-View 3d Pose Estimation](https://mlsanity.com/paper/2404.02041/) (2024)
- [Adaptive Multi-view and Temporal Fusing Transformer for 3D Human Pose Estimation](https://mlsanity.com/paper/2110.05092/) (2021)
- [Multiple View Geometry Transformers for 3D Human Pose Estimation](https://mlsanity.com/paper/2311.10983/) (2023)
- [RTMW: Real-Time Multi-Person 2D and 3D Whole-body Pose Estimation](https://mlsanity.com/paper/2407.08634/) (2024)
- [FLEX: Extrinsic Parameters-free Multi-view 3D Human Motion Reconstruction](https://mlsanity.com/paper/2105.01937/) (2021)
- [HAA4D: Few-Shot Human Atomic Action Recognition via 3D Spatio-Temporal Skeletal Alignment](https://mlsanity.com/paper/2202.07308/) (2022)
- [MPL: Lifting 3D Human Pose from Multi-view 2D Poses](https://mlsanity.com/paper/2408.10805/) (2024)
- [Unconstrained Multi-view Human Pose Estimation with Algebraic Priors](https://mlsanity.com/paper/2604.24312/) (2026)

## Abstract (arXiv)

Monocular 3D human pose estimation (HPE) remains challenging due to depth ambiguity, occlu- sions, and the need for temporal consistency. While multi-view methods provide superior accuracy over monocular approaches, they often require complex setups. We introduce STA-TFM, a transformer-based architecture that combines spatial and temporal information for multi-view pose estimation. The approach leverages DSTformer, a monocular feature extractor, to capture long-range pose dependencies within each view. A fusion transformer then aggregates information across views to produce coherent 3D estimates. To address training data scarcity, we use a data generation pipeline that transforms any existing 3D pose dataset into multi-view setups with controllable parameters. Experiments on various datasets demonstrate that STA-TFM outperforms existing camera-parameter-free multi-view methods. STA-TFM achieves 50.9% and 49.5% reductions in mean per joint position error (MPJPE) and mean per joint velocity error (MPJVE) on the DHP19 dataset. Furthermore, it achieves 6.7% and 7.7% respective reductions on HAA4D, and a 15.2% MPJPE reduction on TotalCapture. STA-TFM handles noisy and missing 2D inputs, supporting potential deployment in healthcare monitoring, athletic assessment, and immersive technologies. Code, training checkpoints, and data are available at https://zenodo.org/records/22832620.