| StarPose: 3D Human Pose Estimation via Spatial-Temporal Autoregressive Diffusion | ✓ Link | 30.5 | | | | | | StarPose (Predicted intrinsics, CPN, T=243, mean N=5) (extra data) | 2025-08-04 |
| FinePOSE: Fine-Grained Prompt-Driven 3D Human Pose Estimation via Diffusion Models | ✓ Link | 31.9 | | | | | | FinePOSE + JPMA (CPN, T=243, H=20, K=10) | 2024-05-08 |
| CHAMP: Conformalized 3D Human Multi-Hypothesis Pose Estimators | | 34.8 | | | | | | CHAMP-Agg | 2024-05-27 |
| L2A: Learning to Accumulate Pose History for Accurate 3D Human Pose Estimation | | 36.0 | | | | | | L2A (L) | 2026-05-09 |
| A Structure-aware and Motion-adaptive Framework for 3D Human Pose Estimation with Mamba | | 36.5 | | | | | | SAMA-L (SHNet, T=351, seq2seq) | 2025-07-26 |
| LangPose: Language-Aligned Motion for Robust 3D Human Pose Estimation | | 36.7 | | | | | | LangPose (SH, T=243) (extra data) | 2024-08-31 |
| HiPART: Hierarchical Pose AutoRegressive Transformer for Occluded 3D Human Pose Estimation | | 36.7 | | | | | | HiPART + MixSTE (f=243) | 2025-03-30 |
| L2A: Learning to Accumulate Pose History for Accurate 3D Human Pose Estimation | | 36.7 | | | | | | L2A (B) | 2026-05-09 |
| DiffPose: Toward More Reliable 3D Pose Estimation | ✓ Link | 36.9 | | | | | | DiffPose GMM-Diff M=5 (CPN, T=243, mean N=5) | 2022-11-30 |
| MotionBERT: A Unified Perspective on Learning Human Motion Representations | ✓ Link | 37.5 | Yes | 243 | No | | SH | MotionBERT (Finetune) | 2022-10-12 |
| TCPFormer: Learning Temporal Correlation with Implicit Pose Proxy for 3D Human Pose Estimation | ✓ Link | 37.9 | | | | | | TCPFormer (SH, T=243, seq2seq) | 2025-01-03 |
| DSR-Net: Learning Dynamic Skeletal Relations for 3D Human Pose Estimation | | 37.9 | | | | | | DSR-Net | 2026-10-10 |
| MotionAGFormer: Enhancing 3D Human Pose Estimation with a Transformer-GCNFormer Network | ✓ Link | 38.4 | Yes | 243 | No | | SH | MotionAGFormer-L | 2023-10-25 |
| MotionAGFormer: Enhancing 3D Human Pose Estimation with a Transformer-GCNFormer Network | ✓ Link | 38.4 | Yes | 243 | No | | SH | MotionAGFormer-B | 2023-10-25 |
| Hourglass Tokenizer for Efficient Transformer-Based 3D Human Pose Estimation | ✓ Link | 39.0 | | | | | | TPC w. MotionBERT (SH, F=243, n=2, f=121, seq2frame) | 2023-11-20 |
| MotionBERT: A Unified Perspective on Learning Human Motion Representations | ✓ Link | 39.2 | Yes | 243 | No | | SH | MotionBERT (Scratch) | 2022-10-12 |
| Diffusion-Based 3D Human Pose Estimation with Multi-Hypothesis Aggregation | ✓ Link | 39.5 | Yes | 243 | No | | CPN | D3DP | 2023-03-21 |
| Disentangled Diffusion-Based 3D Human Pose Estimation with Hierarchical Spatial and Temporal Denoiser | ✓ Link | 39.7 | Yes | 243 | No | | CPN | DDHPose | 2024-03-07 |
| SasMamba: A Lightweight Structure-Aware Stride State Space Model for 3D Human Pose Estimation | | 39.77 | | | | | | SasMamba-large (Stacked Hourglass, T=243, seq2seq) | 2025-11-12 |
| MixSTE: Seq2seq Mixed Spatio-Temporal Encoder for 3D Human Pose Estimation in Video | ✓ Link | 39.8 | Yes | 243 | No | | HRNet | MixSTE (HRNet, T=243) | 2022-03-02 |
| HEMlets Pose: Learning Part-Centric Heatmap Triplets for Accurate 3D Human Pose Estimation | | 39.9 | | 1 | | 27.9 | | HEMlets Pose (H36M+MPII) | 2019-10-26 |
| L2A: Learning to Accumulate Pose History for Accurate 3D Human Pose Estimation | | 40.0 | | | | | | L2A (S) | 2026-05-09 |
| 3D Human Pose Estimation using Spatio-Temporal Networks with Explicit Occlusion Training | | 40.1 | Yes | 128 | No | 30.7 | | Spatio-Temporal Network (T=128) | 2020-04-07 |
| KTPFormer: Kinematics and Trajectory Prior Knowledge-Enhanced Transformer for 3D Human Pose Estimation | ✓ Link | 40.1 | Yes | 243 | No | | CPN | KTPFormer | 2024-03-31 |
| GenHMR: Generative Human Mesh Recovery | | 41.2 | | | | 29.8 | | GenHMR | 2024-12-19 |
| P-STMO: Pre-Trained Spatial Temporal Many-to-One Model for 3D Human Pose Estimation | ✓ Link | 42.1 | Yes | 243 | No | | CPN | P-STMO (N=243) | 2022-03-15 |
| MPM: A Unified 2D-3D Human Pose Representation via Masked Pose Modeling | | 42.3 | | | | | | MPM (CPN, T=243) (extra data) | 2023-06-29 |
| MotionAGFormer: Enhancing 3D Human Pose Estimation with a Transformer-GCNFormer Network | ✓ Link | 42.5 | Yes | 81 | No | | SH | MotionAGFormer-S | 2023-10-25 |
| MHFormer: Multi-Hypothesis Transformer for 3D Human Pose Estimation | ✓ Link | 43.0 | | | | | | MHFormer (CPN, T=351) | 2021-11-24 |
| Anatomy-aware 3D Human Pose Estimation with Bone-based Pose Decomposition | ✓ Link | 44.1 | Yes | 243 | No | | CPN | Anatomy3D | 2020-02-24 |
| 3D Human Pose Estimation with Spatial and Temporal Transformers | ✓ Link | 44.3 | | 81 | | | CPN | PoseFormer (T=81) | 2021-03-18 |
| Improving Robustness and Accuracy via Relative Information Encoding in 3D Human Pose Estimation | ✓ Link | 44.3 | Yes | 243 | No | | CPN | RIE (T=243 CPN) | 2021-07-29 |
| GLA-GCN: Global-local Adaptive Graph Convolutional Network for 3D Human Pose Estimation from Monocular Video | ✓ Link | 44.4 | | | | | | GLA-GCN (CPN, T=243) | 2023-07-12 |
| Attention Mechanism Exploits Temporal Contexts: Real-Time 3D Human Pose Reconstruction | ✓ Link | 45.1 | Yes | 243 | No | | CPN | Attention3DHumanPose | 2020-06-01 |
| MotionAGFormer: Enhancing 3D Human Pose Estimation with a Transformer-GCNFormer Network | ✓ Link | 45.1 | Yes | 27 | No | | SH | MotionAGFormer-XS | 2023-10-25 |
| PoseFormerV2: Exploring Frequency Domain for Efficient and Robust 3D Human Pose Estimation | ✓ Link | 45.2 | | | | | | PoseFormerV2 (f=27, n=27, sequence=243) | 2023-03-30 |
| Trajectory Space Factorization for Deep Video-Based 3D Human Pose Estimation | ✓ Link | 46.6 | Yes | 50 | No | | | Trajectory Space Factorization (50 frames) | 2019-08-22 |
| 3D human pose estimation in video with temporal convolutions and semi-supervised training | ✓ Link | 46.8 | Yes | 243 | No | | CPN | VideoPose3D (T=243) | 2018-11-28 |
| Sampling is Matter: Point-guided 3D Human Mesh Reconstruction | ✓ Link | 48.3 | | | | 32.9 | | PointHMR | 2023-04-19 |
| SRNet: Improving Generalization in 3D Human Pose Estimation with a Split-and-Recombine Approach | ✓ Link | 49.9 | No | 1 | No | | | SRNET | 2020-07-18 |
| PoseAug: A Differentiable Pose Augmentation Framework for 3D Human Pose Estimation | ✓ Link | 50.2 | | | | 39.1 | | HR-Net+VPose+PoseAug | 2021-05-06 |
| Cascaded deep monocular 3D human pose estimation with evolutionary training data | ✓ Link | 50.9 | No | 1 | No | | | TAG-Net | 2020-06-14 |
| Occlusion Resilient 3D Human Pose Estimation | | 51.0 | | | | | | Occlusion Resilient 3D Human Pose Estimation (Ours) | 2024-02-16 |
| Skeleton-in-Context: Unified Skeleton Sequence Modeling with In-Context Learning | ✓ Link | 51.6 | | | | | | Static-SiC | 2023-12-06 |
| Predicting Camera Viewpoint Improves Cross-dataset Generalization for 3D Human Pose Estimation | | 52.0 | No | 1 | No | | | cross-dataset-evaluation | 2020-04-07 |
| Learning 3D Human Pose from Structure and Motion | ✓ Link | 52.1 | Yes | 20 | No | | | TP-Net | 2017-11-25 |
| Generating Multiple Hypotheses for 3D Human Pose Estimation with Mixture Density Network | ✓ Link | 52.7 | No | 1 | No | | | Multimodal Mixture Density Networks | 2019-04-11 |
| Semantic Graph Convolutional Networks for 3D Human Pose Regression | ✓ Link | 57.6 | No | 1 | No | | | SemGCN | 2019-04-06 |
| Monocular 3D Human Pose Estimation by Generation and Ordinal Ranking | ✓ Link | 58.0 | No | 1 | No | | | MultiPoseNet | 2019-04-02 |
| Monocular Total Capture: Posing Face, Body, and Hands in the Wild | ✓ Link | 58.3 | NO | 1 | No | | | Monocular Total Capture | 2018-12-04 |
| A simple yet effective baseline for 3d human pose estimation | ✓ Link | 62.9 | No | 1 | No | | | SIM (SH detections FT) (MA) | 2017-05-08 |
| Exploiting temporal context for 3D human pose estimation in the wild | ✓ Link | 63.3 | | | | | | Bundle Adjustment (GTi) | 2019-05-10 |
| XNect: Real-time Multi-Person 3D Motion Capture with a Single RGB Camera | ✓ Link | 63.6 | No | 1 | No | | | SelecSLS | 2019-07-01 |
| Towards 3D Human Pose Estimation in the Wild: a Weakly-supervised Approach | ✓ Link | 64.9 | No | 1 | No | | | Weakly Supervised Transfer Learning | 2017-04-08 |
| VIBE: Video Inference for Human Body Pose and Shape Estimation | ✓ Link | 65.6 | Yes | 16 | No | | | VIBE | 2019-12-11 |
| Convolutional Mesh Regression for Single-Image Human Shape Reconstruction | ✓ Link | 74.7 | No | 1 | No | | | GraphCMR | 2019-05-08 |
| Lifting from the Deep: Convolutional 3D Pose Estimation from a Single Image | ✓ Link | 88.39 | No | 1 | No | | | Projected-pose belief maps + 2D fusion layers | 2017-01-01 |
| RepNet: Weakly Supervised Training of an Adversarial Reprojection Network for 3D Human Pose Estimation | ✓ Link | 89.9 | No | 1 | No | | | RepNet | 2019-02-26 |
| A Dual-Source Approach for 3D Human Pose Estimation from a Single Image | | 97.39 | No | 1 | No | | | Dual-source approach | 2017-05-08 |
| Sparseness Meets Deepness: 3D Human Pose Estimation from Monocular Video | ✓ Link | 113.01 | Yes | 300 | No | | | Sparseness Meets Deepness | 2015-11-30 |
| End-to-end Recovery of Human Shape and Pose | ✓ Link | | No | 1 | No | | | HMR | 2017-12-18 |
| Neural Body Fitting: Unifying Deep Learning and Model-Based Human Pose and Shape Estimation | ✓ Link | | No | 1 | No | | | Neural Body Fitting
(NBF) | 2018-08-17 |
| Unite the People: Closing the Loop Between 3D and 2D Human Representations | ✓ Link | | No | 1 | No | | | SMPLify
(dense) | 2017-01-10 |
| Ordinal Depth Supervision for 3D Human Pose Estimation | ✓ Link | | No | 1 | No | | | Ordinal Depth Supervision | 2018-05-10 |
| 3D Human Pose Estimation in the Wild by Adversarial Learning | | | No | 1 | No | | | Adversarial Learning | 2018-03-26 |
| Camera Distance-aware Top-down Approach for 3D Multi-person Pose Estimation from a Single RGB Image | ✓ Link | | No | 1 | No | | | Moon et. al. | 2019-07-26 |
| PoseAug: A Differentiable Pose Augmentation Framework for 3D Human Pose Estimation | ✓ Link | | No | 1 | No | | | PoseAug | 2021-05-06 |
| HEMlets Pose: Learning Part-Centric Heatmap Triplets for Accurate 3D Human Pose Estimation | | | No | 1 | No | | | HEMlets Pose | 2019-10-26 |
| Ray3D: ray-based 3D human pose estimation for monocular absolute 3D localization | ✓ Link | | Yes | 9 | No | | | Ray3D | 2022-03-22 |
| Exploiting temporal context for 3D human pose estimation in the wild | ✓ Link | | Yes | 190 | No | | | Bundle Adjustment | 2019-05-10 |
| Neural Body Fitting: Unifying Deep Learning and Model-Based Human Pose and Shape Estimation | ✓ Link | | | | | 59.9 | | Neural Body Fitting (NBF) | 2018-08-17 |
| Unite the People: Closing the Loop Between 3D and 2D Human Representations | ✓ Link | | | | | 80.7 | | SMPLify (dense) | 2017-01-10 |