# Field Converter: Geometry-Initialized Temporal Residual Refinement for World-Grounded Player Pose Estimation from Soccer Broadcasts

arXiv: https://arxiv.org/abs/2609.10498 · HTML: https://mlsanity.com/paper/2609.10498/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 23 of its 32 references that are in the map.
- First posted on arXiv 9 Sep 2026.

## Builds on (most cited first, 23 of 23)

- [ORB-SLAM: a Versatile and Accurate Monocular SLAM System](https://mlsanity.com/paper/1502.00956/) (2015)
- [3D human pose estimation in video with temporal convolutions and semi-supervised training](https://mlsanity.com/paper/1811.11742/) (2018)
- [Humans in 4D: Reconstructing and Tracking Humans with Transformers](https://mlsanity.com/paper/2305.20091/) (2023)
- [HybrIK: A Hybrid Analytical-Neural Inverse Kinematics Solution for 3D Human Pose and Shape Estimation](https://mlsanity.com/paper/2011.14672/) (2020)
- [WHAM: Reconstructing World-grounded Humans with Accurate 3D Motion](https://mlsanity.com/paper/2312.07531/) (2023)
- [MotionBERT: A Unified Perspective on Learning Human Motion Representations](https://mlsanity.com/paper/2210.06551/) (2022)
- [MixSTE: Seq2seq Mixed Spatio-Temporal Encoder for 3D Human Pose Estimation in Video](https://mlsanity.com/paper/2203.00859/) (2022)
- [GLAMR: Global Occlusion-Aware Human Mesh Recovery with Dynamic Cameras](https://mlsanity.com/paper/2112.01524/) (2021)
- [Decoupling Human and Camera Motion from Videos in the Wild](https://mlsanity.com/paper/2302.12827/) (2023)
- [World-Grounded Human Motion Recovery via Gravity-View Coordinates](https://mlsanity.com/paper/2409.06662/) (2024)
- [TRAM: Global Trajectory and Motion of 3D Humans from in-the-wild Videos](https://mlsanity.com/paper/2403.17346/) (2024)
- [Multi-HMR: Multi-Person Whole-Body Human Mesh Recovery in a Single Shot](https://mlsanity.com/paper/2402.14654/) (2024)
- [SAM 3D Body: Robust Full-Body Human Mesh Recovery](https://mlsanity.com/paper/2602.15989/) (2026)
- [MotionAGFormer: Enhancing 3D Human Pose Estimation with a Transformer-GCNFormer Network](https://mlsanity.com/paper/2310.16288/) (2023)
- [PACE: Human and Camera Motion Estimation from in-the-wild Videos](https://mlsanity.com/paper/2310.13768/) (2023)
- [Ray3D: ray-based 3D human pose estimation for monocular absolute 3D localization](https://mlsanity.com/paper/2203.11471/) (2022)
- [TVCalib: Camera Calibration for Sports Field Registration in Soccer](https://mlsanity.com/paper/2207.11709/) (2022)
- [AutoSoccerPose: Automated 3D posture Analysis of Soccer Shot Movements](https://mlsanity.com/paper/2405.12070/) (2024)
- [PnLCalib: Sports Field Registration via Points and Lines Optimization](https://mlsanity.com/paper/2404.08401/) (2024)
- [WorldPose: A World Cup Dataset for Global 3D Human Pose Estimation](https://mlsanity.com/paper/2501.02771/) (2025)
- [AthletePose3D: A Benchmark Dataset for 3D Human Pose Estimation and Kinematic Validation in Athletic Movements](https://mlsanity.com/paper/2503.07499/) (2025)
- [Monocular 3D Human Pose Estimation for Sports Broadcasts using Partial Sports Field Registration](https://mlsanity.com/paper/2304.04437/) (2023)
- [ProxyCap: Real-time Monocular Full-body Capture in World Space via Human-Centric Proxy-to-Motion Learning](https://mlsanity.com/paper/2307.01200/) (2023)

## Abstract (arXiv)

Recovering 3D human pose from monocular sports broadcasts remains challenging when players must be localized in a shared metric world coordinate system rather than only reconstructed relative to their own body. We introduce Field Converter, a geometry-initialized temporal residual framework for world-grounded 3D player pose estimation from calibrated soccer broadcasts. Our method first uses camera and pitch geometry to initialize the player root through ray-ground intersection, then predicts a temporal residual correction from pose, image, camera, and geometric cues. On match-disjoint evaluation sequences, residual refinement reduces root error from 49cm with geometry alone to 14cm with a frame-wise MLP and 10cm with a TCN, while a Transformer achieves a comparable 11cm. The resulting world-space MPJPE reaches 13.2cm, and ablations show that residual prediction clearly outperforms direct global-root regression while temporal context matters more than the specific temporal backbone. Failure analysis further identifies airborne motion as the main limitation of the ground-based geometric initialization.