Papers › Out-of-Domain Human Mesh Reconstruction via Dynamic Bilevel Online Adaptation

Out-of-Domain Human Mesh Reconstruction via Dynamic Bilevel Online Adaptation

7 Nov 2021arXiv:2111.04017archive 2025-07-28

Shanyan Guan, Jingwei Xu, Michelle Z. He, Yunbo Wang, Bingbing Ni, Xiaokang Yang

We consider a new problem of adapting a human mesh reconstruction model to out-of-domain streaming videos, where performance of existing SMPL-based models are significantly affected by the distribution shift represented by different camera parameters, bone lengths, backgrounds, and occlusions. We tackle this problem through online adaptation, gradually correcting the model bias during testing. There are two main challenges: First, the lack of 3D annotations increases the training difficulty and results in 3D ambiguities. Second, non-stationary data distribution makes it difficult to strike a balance between fitting regular frames and hard samples with severe occlusions or dramatic changes. To this end, we propose the Dynamic Bilevel Online Adaptation algorithm (DynaBOA). It first introduces the temporal constraints to compensate for the unavailable 3D annotations, and leverages a bilevel optimization procedure to address the conflicts between multi-objectives. DynaBOA provides additional 3D guidance by co-training with similar source examples retrieved efficiently despite the distribution shift. Furthermore, it can adaptively adjust the number of optimization steps on individual frames to fully fit hard samples and avoid overfitting regular frames. DynaBOA achieves state-of-the-art results on three out-of-domain human mesh reconstruction benchmarks.

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

syguan96/dynaboa officialmentioned in papermentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

3D Absolute Human Pose Estimation3D Human Pose EstimationBilevel Optimization

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
3D Absolute Human Pose Estimation Surreal DynaBOA MPJPE 55.2 #1 of 1 Archive leaderboard report
3D Human Pose Estimation 3DPW DynaBOA (w/ 2D GT) MPJPE 65.5 #21 of 119 Archive leaderboard report
3D Human Pose Estimation 3DPW DynaBOA (w/ 2D GT) MPVPE 82 #21 of 119 Archive leaderboard report
3D Human Pose Estimation 3DPW DynaBOA (w/ 2D GT) PA-MPJPE 40.4 #21 of 119 Archive leaderboard report
3D Human Pose Estimation MPI-INF-3DHP DynaBOA AUC 43.1 #69 of 108 Archive leaderboard report
3D Human Pose Estimation MPI-INF-3DHP DynaBOA MPJPE 101.5 #69 of 108 Archive leaderboard report
3D Human Pose Estimation MPI-INF-3DHP DynaBOA PA-MPJPE 66.1 #69 of 108 Archive leaderboard report
3D Human Pose Estimation MPI-INF-3DHP DynaBOA PCK 79.5 #69 of 108 Archive leaderboard report
3D Human Pose Estimation Surreal DynaBOA PA-MPJPE 34 #6 of 6 Archive leaderboard report
3D Human Pose Estimation Surreal DynaBOA PVE 70.7 #6 of 6 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections