# WildHSR: Metric Feed-Forward 4D People-Scene Reconstruction from a 3D Foundation Model

arXiv: https://arxiv.org/abs/2609.29106 · HTML: https://mlsanity.com/paper/2609.29106/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 38 of its 43 references that are in the map.
- First posted on arXiv 24 Sep 2026.

## Builds on (most cited first, 30 of 38)

- [Emerging Properties in Self-Supervised Vision Transformers](https://mlsanity.com/paper/2104.14294/) (2021)
- [Sinkhorn Distances: Lightspeed Computation of Optimal Transportation Distances](https://mlsanity.com/paper/1306.0895/) (2013)
- [Expressive Body Capture: 3D Hands, Face, and Body from a Single Image](https://mlsanity.com/paper/1904.05866/) (2019)
- [DUSt3R: Geometric 3D Vision Made Easy](https://mlsanity.com/paper/2312.14132/) (2023)
- [VGGT: Visual Geometry Grounded Transformer](https://mlsanity.com/paper/2503.11651/) (2025)
- [Understanding intermediate layers using linear classifier probes](https://mlsanity.com/paper/1610.01644/) (2016)
- [Self-training with Noisy Student improves ImageNet classification](https://mlsanity.com/paper/1911.04252/) (2019)
- [Depth Map Prediction from a Single Image using a Multi-Scale Deep Network](https://mlsanity.com/paper/1406.2283/) (2014)
- [Continuous 3D Perception Model with Persistent State](https://mlsanity.com/paper/2501.12387/) (2025)
- [MonST3R: A Simple Approach for Estimating Geometry in the Presence of Motion](https://mlsanity.com/paper/2410.03825/) (2024)
- [Humans in 4D: Reconstructing and Tracking Humans with Transformers](https://mlsanity.com/paper/2305.20091/) (2023)
- [MapAnything: Universal Feed-Forward Metric 3D Reconstruction](https://mlsanity.com/paper/2509.13414/) (2025)
- [WHAM: Reconstructing World-grounded Humans with Accurate 3D Motion](https://mlsanity.com/paper/2312.07531/) (2023)
- [Capturing and Inferring Dense Full-Body Human-Scene Contact](https://mlsanity.com/paper/2206.09553/) (2022)
- [Probing the 3D Awareness of Visual Foundation Models](https://mlsanity.com/paper/2404.08636/) (2024)
- [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)
- [EMDB: The Electromagnetic Database of Global 3D Human Pose and Shape in the Wild](https://mlsanity.com/paper/2308.16894/) (2023)
- [TRACE: 5D Temporal Regression of Avatars with Dynamic Cameras in 3D Environments](https://mlsanity.com/paper/2306.02850/) (2023)
- [Easi3R: Estimating Disentangled Motion from DUSt3R Without Training](https://mlsanity.com/paper/2503.24391/) (2025)
- [PromptHMR: Promptable Human Mesh Recovery](https://mlsanity.com/paper/2504.06397/) (2025)
- [ViTPose++: Vision Transformer for Generic Body Pose Estimation](https://mlsanity.com/paper/2212.04246/) (2022)
- [Human3R: Everyone Everywhere All at Once](https://mlsanity.com/paper/2510.06219/) (2025)
- [PACE: Human and Camera Motion Estimation from in-the-wild Videos](https://mlsanity.com/paper/2310.13768/) (2023)
- [Reconstructing People, Places, and Cameras](https://mlsanity.com/paper/2412.17806/) (2024)
- [Joint Optimization for 4D Human-Scene Reconstruction in the Wild](https://mlsanity.com/paper/2501.02158/) (2025)
- [AMB3R: Accurate Feed-forward Metric-scale 3D Reconstruction with Backend](https://mlsanity.com/paper/2511.20343/) (2025)

## Abstract (arXiv)

3D foundation models recover video cameras and geometry in one forward pass, but some of the strongest are up to scale. Joint people-scene reconstruction then requires two missing outputs: metric scale and persistent person identity. We ask whether one up-to-scale foundation representation can support both through lightweight adaptation. Exact metric labels are scarce, but unlabeled in-the-wild video is abundant. We use people in curated web video to initialise the solution: a posed metric body and 2D keypoints give an approximate, closed-form scale pseudo-label. These pseudo-labels pretrain a Scale Readout, which is then fine-tuned together with a lightweight adapter using exact metric supervision from standard real-video training splits. At inference the head predicts metric scale from foundation-model tokens, without the ruler or its teachers. For person identity, we probe the pretrained foundation model alone and find evidence that its intermediate query-key features encode person correspondence across frames. In most evaluated moving-person clips, a mid-layer token prefers that person over the vacated location and other people. A tiny projection reads this correspondence; together with metric pelvis motion and proposal confidence, it drives dustbin-aware Sinkhorn association of per-frame bodies. WildHSR combines both readouts to reconstruct metric cameras, scene and people from monocular video. Each window is predicted feed-forward; analytic association and Sim(3) composition connect windows. On EMDB-2, WildHSR is the first feed-forward method in the published comparison to beat the best optimization-based WA-MPJPE and RTE while leading feed-forward methods on all three world-frame metrics. On RICH, it leads feed-forward people-and-scene methods on WA-MPJPE and W-MPJPE. The complete pipeline runs at 10.1 fps on one GPU.