# Human Mesh Recovery on 3DPW standard (test)

Source: [Human Mesh Recovery on 3DPW standard (test)](<https://www.sota2.com/research/sota/human-mesh-recovery-on-3dpw-standard-test>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T05:59:46.119Z

Evaluation of Human Mesh Recovery performance on the standard test split of the 3DPW dataset, reporting metrics including MPJPE and RE.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Human Mesh Recovery |
| Dataset (including any stated split) | 3DPW standard (test) |
| Metric | RE |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 35 of 35 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [RE](<https://www.sota2.com/research/sota/human-mesh-recovery-on-3dpw-standard-test.md?metric=RE>): Lower is better; unit: SCALAR.

- [MPJPE](<https://www.sota2.com/research/sota/human-mesh-recovery-on-3dpw-standard-test.md?metric=MPJPE>): Lower is better; unit: SCALAR.

- [PA-MPJPE](<https://www.sota2.com/research/sota/human-mesh-recovery-on-3dpw-standard-test.md?metric=PA-MPJPE>): Lower is better; unit: SCALAR.

## Recorded results

| Method | RE | Result date | Report paper |
| --- | --- | --- | --- |
| n-quantized-best-of-M | 55.6 | Not recorded | 68fa68aa1e209131cea1d95a |
| n-quantized-best-of-M | 56.6 | Not recorded | 68fa68aa1e209131cea1d95a |
| n-quantized-best-of-M | 57.1 | Not recorded | 68fa68aa1e209131cea1d95a |
| SPIN | 59.3 | Not recorded | 68fa68aa1e209131cea1d95a |
| n-quantized-best-of-M | 59.9 | Not recorded | 68fa68aa1e209131cea1d95a |
| SMPL-MDN | 60.1 | Not recorded | 68fa68aa1e209131cea1d95a |
| SMPL-CVAE | 60.3 | Not recorded | 68fa68aa1e209131cea1d95a |
| SMPL-CVAE | 60.5 | Not recorded | 68fa68aa1e209131cea1d95a |
| SMPL-MDN | 60.7 | Not recorded | 68fa68aa1e209131cea1d95a |
| SMPL-CVAE | 60.7 | Not recorded | 68fa68aa1e209131cea1d95a |
| SMPL-MDN | 61.2 | Not recorded | 68fa68aa1e209131cea1d95a |
| SMPL-CVAE | 61.4 | Not recorded | 68fa68aa1e209131cea1d95a |
| SMPL-MDN | 64.7 | Not recorded | 68fa68aa1e209131cea1d95a |
| GraphCMR | 70.2 | Not recorded | 68fa68aa1e209131cea1d95a |
| n-quantized-best-of-M | 72.1 | Not recorded | 68fa68aa1e209131cea1d95a |
| n-quantized-best-of-M | 73.7 | Not recorded | 68fa68aa1e209131cea1d95a |
| n-quantized-best-of-M | 74.4 | Not recorded | 68fa68aa1e209131cea1d95a |
| SMPL-MDN | 76.6 | Not recorded | 68fa68aa1e209131cea1d95a |
| n-quantized-best-of-M | 78.5 | Not recorded | 68fa68aa1e209131cea1d95a |
| SMPL-CVAE | 79.8 | Not recorded | 68fa68aa1e209131cea1d95a |
| SMPL-CVAE | 79.8 | Not recorded | 68fa68aa1e209131cea1d95a |
| SMPL-CVAE | 79.9 | Not recorded | 68fa68aa1e209131cea1d95a |
| SMPL-CVAE | 80.2 | Not recorded | 68fa68aa1e209131cea1d95a |
| SMPL-MDN | 80.7 | Not recorded | 68fa68aa1e209131cea1d95a |
| HMR | 81.3 | Not recorded | 68fa68aa1e209131cea1d95a |
| SMPL-MDN | 82.8 | Not recorded | 68fa68aa1e209131cea1d95a |
| SMPL-MDN | 83 | Not recorded | 68fa68aa1e209131cea1d95a |

### n-quantized-best-of-M

RE: 55.6. Result ID: 698d7d6b8b517eb43321e7a9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

Recorded evaluation configuration: {"Quantization n":25}. This record does not assert independent reproduction.

### n-quantized-best-of-M

RE: 56.6. Result ID: 698d7d6b8b517eb43321e7a8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

Recorded evaluation configuration: {"Quantization n":10}. This record does not assert independent reproduction.

### n-quantized-best-of-M

RE: 57.1. Result ID: 698d7d6b8b517eb43321e7a7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

Recorded evaluation configuration: {"Quantization n":5}. This record does not assert independent reproduction.

### SPIN

RE: 59.3. Result ID: 698d7d6b8b517eb43321e79d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa667a1e209131cea106f3.md>)

[Implementation](<https://seas.upenn.edu/~nkolot/projects/spin>)

Recorded evaluation configuration: {"Quantization n":1}. This record does not assert independent reproduction.

### n-quantized-best-of-M

RE: 59.9. Result ID: 698d7d6b8b517eb43321e7a6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

Recorded evaluation configuration: {"Quantization n":1}. This record does not assert independent reproduction.

### SMPL-MDN

RE: 60.1. Result ID: 698d7d6b8b517eb43321e7a1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65a51e209131cea0b8e8.md>)

[Implementation](<https://github.com/chaneyddtt/Generating-Multiple-Hypotheses-for-3D-Human-Pose-Estimation-with-Mixture-Density-Network>)

Recorded evaluation configuration: {"Quantization n":25}. This record does not assert independent reproduction.

### SMPL-CVAE

RE: 60.3. Result ID: 698d7d6b8b517eb43321e7a5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

Recorded evaluation configuration: {"Quantization n":25}. This record does not assert independent reproduction.

### SMPL-CVAE

RE: 60.5. Result ID: 698d7d6b8b517eb43321e7a4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

Recorded evaluation configuration: {"Quantization n":10}. This record does not assert independent reproduction.

### SMPL-MDN

RE: 60.7. Result ID: 698d7d6b8b517eb43321e7a0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65a51e209131cea0b8e8.md>)

[Implementation](<https://github.com/chaneyddtt/Generating-Multiple-Hypotheses-for-3D-Human-Pose-Estimation-with-Mixture-Density-Network>)

Recorded evaluation configuration: {"Quantization n":10}. This record does not assert independent reproduction.

### SMPL-CVAE

RE: 60.7. Result ID: 698d7d6b8b517eb43321e7a3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

Recorded evaluation configuration: {"Quantization n":5}. This record does not assert independent reproduction.

### SMPL-MDN

RE: 61.2. Result ID: 698d7d6b8b517eb43321e79f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65a51e209131cea0b8e8.md>)

[Implementation](<https://github.com/chaneyddtt/Generating-Multiple-Hypotheses-for-3D-Human-Pose-Estimation-with-Mixture-Density-Network>)

Recorded evaluation configuration: {"Quantization n":5}. This record does not assert independent reproduction.

### SMPL-CVAE

RE: 61.4. Result ID: 698d7d6b8b517eb43321e7a2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

Recorded evaluation configuration: {"Quantization n":1}. This record does not assert independent reproduction.

### SMPL-MDN

RE: 64.7. Result ID: 698d7d6b8b517eb43321e79e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65a51e209131cea0b8e8.md>)

[Implementation](<https://github.com/chaneyddtt/Generating-Multiple-Hypotheses-for-3D-Human-Pose-Estimation-with-Mixture-Density-Network>)

Recorded evaluation configuration: {"Quantization n":1}. This record does not assert independent reproduction.

### GraphCMR

RE: 70.2. Result ID: 698d7d6b8b517eb43321e79c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65c11e209131cea0c393.md>)

[Implementation](<https://www.sota2.com/seas.upenn.edu/~nkolot/projects/cmr>)

Recorded evaluation configuration: {"Quantization n":1}. This record does not assert independent reproduction.

### n-quantized-best-of-M

RE: 72.1. Result ID: 698d7d6b8b517eb43321e7c1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

Recorded evaluation configuration: {"Quantization n":25}. This record does not assert independent reproduction.

### n-quantized-best-of-M

RE: 73.7. Result ID: 698d7d6b8b517eb43321e7c0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

Recorded evaluation configuration: {"Quantization n":10}. This record does not assert independent reproduction.

### n-quantized-best-of-M

RE: 74.4. Result ID: 698d7d6b8b517eb43321e7bf.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

Recorded evaluation configuration: {"Quantization n":5}. This record does not assert independent reproduction.

### SMPL-MDN

RE: 76.6. Result ID: 698d7d6b8b517eb43321e7b9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65a51e209131cea0b8e8.md>)

[Implementation](<https://github.com/chaneyddtt/Generating-Multiple-Hypotheses-for-3D-Human-Pose-Estimation-with-Mixture-Density-Network>)

Recorded evaluation configuration: {"Quantization n":25}. This record does not assert independent reproduction.

### n-quantized-best-of-M

RE: 78.5. Result ID: 698d7d6b8b517eb43321e7be.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

Recorded evaluation configuration: {"Quantization n":1}. This record does not assert independent reproduction.

### SMPL-CVAE

RE: 79.8. Result ID: 698d7d6b8b517eb43321e7bc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

Recorded evaluation configuration: {"Quantization n":10}. This record does not assert independent reproduction.

### SMPL-CVAE

RE: 79.8. Result ID: 698d7d6b8b517eb43321e7bd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

Recorded evaluation configuration: {"Quantization n":25}. This record does not assert independent reproduction.

### SMPL-CVAE

RE: 79.9. Result ID: 698d7d6b8b517eb43321e7bb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

Recorded evaluation configuration: {"Quantization n":5}. This record does not assert independent reproduction.

### SMPL-CVAE

RE: 80.2. Result ID: 698d7d6b8b517eb43321e7ba.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

Recorded evaluation configuration: {"Quantization n":1}. This record does not assert independent reproduction.

### SMPL-MDN

RE: 80.7. Result ID: 698d7d6b8b517eb43321e7b8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65a51e209131cea0b8e8.md>)

[Implementation](<https://github.com/chaneyddtt/Generating-Multiple-Hypotheses-for-3D-Human-Pose-Estimation-with-Mixture-Density-Network>)

Recorded evaluation configuration: {"Quantization n":10}. This record does not assert independent reproduction.

### HMR

RE: 81.3. Result ID: 698d7d6b8b517eb43321e79b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa647e1e209131cea04d88.md>)

[Implementation](<https://akanazawa.github.io/hmr/>)

Recorded evaluation configuration: {"Quantization n":1}. This record does not assert independent reproduction.

### SMPL-MDN

RE: 82.8. Result ID: 698d7d6b8b517eb43321e7b6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65a51e209131cea0b8e8.md>)

[Implementation](<https://github.com/chaneyddtt/Generating-Multiple-Hypotheses-for-3D-Human-Pose-Estimation-with-Mixture-Density-Network>)

Recorded evaluation configuration: {"Quantization n":1}. This record does not assert independent reproduction.

### SMPL-MDN

RE: 83. Result ID: 698d7d6b8b517eb43321e7b7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa68aa1e209131cea1d95a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65a51e209131cea0b8e8.md>)

[Implementation](<https://github.com/chaneyddtt/Generating-Multiple-Hypotheses-for-3D-Human-Pose-Estimation-with-Mixture-Density-Network>)

Recorded evaluation configuration: {"Quantization n":5}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.