| PRNet | 1.960 (±0.731) | 2.445 (±0.570) | 1.856 (±0.607) | 1.868 (±0.510) | 2.032 | Joint 3D Face Reconstruction and Dense Alignment with Position Map Regression Network | |
| HiFace-f | 1.360 (±0.395) | 1.399 (±0.388) | 1.489 (±0.436) | 0.985 (±0.237) | 1.308 | HiFace: High-Fidelity 3D Face Reconstruction by Learning Static and Dynamic Details | - |
| Deep3D | 1.528 (±0.517) | 2.074 (±0.486) | 1.411 (±0.395) | 1.749 (±0.343) | 1.657 | Accurate 3D Face Reconstruction with Weakly-Supervised Learning: From Single Image to Image Set | |
| PSL | 1.469 (±0.495) | 2.454 (±0.608) | 1.820 (±0.557) | 1.685 (±0.475) | 1.857 | A Perceptual Shape Loss for Monocular 3D Face Reconstruction | - |
| EMOCA-f | 1.599 (±0.563) | 2.606 (±0.686) | 2.948 (±1.292) | 2.455 (±0.636) | 2.402 | EMOCA: Emotion Driven Monocular Face Capture and Animation | |
| HRN | 1.038 (±0.322) | 1.906 (±0.479) | 1.285 (±0.528) | 1.642 (±0.310) | 1.468 | A Hierarchical Representation Network for Accurate and Detailed Face Reconstruction from In-The-Wild Images | |
| 3DDFA-v3 | 1.073 (±0.316) | 1.861 (±0.421) | 1.209 (±0.369) | 1.621 (±0.312) | 1.441 | 3D Face Reconstruction with the Geometric Guidance of Facial Part Segmentation | |
| SynergyNet | 1.662 (±0.627) | 2.638 (±0.719) | 1.725 (±0.533) | 2.008 (±0.526) | 2.008 | Synergy between 3DMM and 3D Landmarks for Accurate 3D Facial Geometry | |
| ExpNet | 1.842 (±0.609) | 3.393 (±1.076) | 2.160 (±0.448) | 2.508 (±0.491) | 2.476 | ExpNet: Landmark-Free, Deep, 3D Facial Expressions | |
| SADRNet | 2.010 (±0.715) | 2.490 (±0.566) | 1.560 (±0.462) | 1.771 (±0.521) | 1.958 | SADRNet: Self-Aligned Dual Face Regression Networks for Robust 3D Dense Face Alignment and Reconstruction | |
| MGCNet | 1.665 (±0.644) | 2.248 (±0.508) | 1.409 (±0.418) | 1.827 (±0.383) | 1.787 | Self-Supervised Monocular 3D Face Reconstruction by Occlusion-Aware Multi-view Geometry Consistency | |
| AlbedoGAN | 1.112 (±0.278) | 2.142 (±0.554) | 2.218 (±0.952) | 1.576 (±0.338) | 1.762 | Towards Realistic Generative 3D Face Models | |
| EMOCA-c | 1.548 (±0.590) | 2.448 (±0.708) | 2.636 (±1.284) | 1.867 (±0.554) | 2.125 | EMOCA: Emotion Driven Monocular Face Capture and Animation | |
| DECA-f | 1.555 (±0.822) | 2.519 (±0.718) | 2.684 (±1.041) | 2.286 (±1.103) | 2.261 | Learning an Animatable Detailed 3D Face Model from In-The-Wild Images | |
| 3DDFA-v2 | 1.781 (±0.636) | 2.465 (±0.622) | 1.642 (±0.501) | 1.883 (±0.499) | 1.943 | Towards Fast, Accurate and Stable 3D Dense Face Alignment | |
| DECA-c | 1.630 (±1.135) | 2.423 (±0.720) | 2.472 (±1.079) | 1.903 (±1.050) | 2.107 | Learning an Animatable Detailed 3D Face Model from In-The-Wild Images | |
| HiFace-c | 1.4392 (±0.429) | 1.427 (±0.400) | 1.505 (±0.454) | 0.992 (±0.246) | 1.341 | HiFace: High-Fidelity 3D Face Reconstruction by Learning Static and Dynamic Details | - |
| MICA | 1.109 (±0.325) | 2.379 (±0.675) | 3.567 (±1.212) | 1.525 (±0.322) | 2.145 | Towards Metrical Reconstruction of Human Faces | |
| RingNet | 2.027 (±0.710) | 3.081 (±0.950) | 1.994 (±0.604) | 1.921 (±0.451) | 2.256 | Learning to Regress 3D Face Shape and Expression from an Image without 3D Supervision | |