Head Pose Estimation on AFLW2000 (Euler and Vector Metrics)
3.03Euler Yaw ErrorLiu et al.
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Liu et al.Own implementation=false, Wider rotation range=false, Training dataset=300W-LP2023.09 | 3.03 | 5.06 | 3.68 | 3.93 | — | — | — | — | |
| RankPoseOwn implementation=true, Wider rotation range=false, Training dataset=300W-LP2023.09 | 3.26 | 4.72 | 3.23 | 3.74 | 4.4 | 4.42 | 5.08 | 4.63 | |
| 6DRepNetOwn implementation=false, Wider rotation range=false, Training dataset=300W-LP2023.09 | 3.27 | 4.58 | 2.98 | 3.61 | 4.33 | 4.17 | 5.06 | 4.52 | |
| MNNOwn implementation=false, Wider rotation range=false, Training dataset=300W-LP2023.09 | 3.34 | 4.69 | 3.48 | 3.83 | — | — | — | — | |
| img2poseOwn implementation=true, Wider rotation range=false, Training dataset=300W-LP2023.09 | 3.42 | 5.03 | 3.28 | 3.91 | 6 | 5.2 | 6.55 | 5.92 | |
| 6DRepNet360Own implementation=false, Wider rotation range=true, Training dataset=CMU+300W-LP2023.09 | 3.73 | 5.52 | 3.53 | 4.26 | 5.18 | 4.7 | 6.04 | 5.31 | |
| FDNOwn implementation=false, Wider rotation range=false, Training dataset=300W-LP2023.09 | 3.78 | 5.61 | 3.88 | 4.42 | — | — | — | — | |
| QuatNetOwn implementation=false, Wider rotation range=false, Training dataset=300W-LP2023.09 | 3.97 | 5.62 | 3.92 | 4.5 | — | — | — | — | |
| MFDNetOwn implementation=false, Wider rotation range=false, Training dataset=300W-LP2023.09 | 4.3 | 5.16 | 3.69 | 4.38 | — | — | — | — | |
| TriNetOwn implementation=true, Wider rotation range=false, Training dataset=300W-LP2023.09 | 4.36 | 5.81 | 4.51 | 4.89 | 6.16 | 5.95 | 6.82 | 6.31 | |
| DDD-PoseOwn implementation=false, Wider rotation range=false, Training dataset=300W-LP2023.09 | 4.38 | 4.85 | 3.44 | 4.22 | — | — | — | — | |
| WHENet-VOwn implementation=false, Wider rotation range=false, Training dataset=300W-LP2023.09 | 4.44 | 5.75 | 4.31 | 4.83 | — | — | — | — | |
| HPEOwn implementation=false, Wider rotation range=false, Training dataset=300W-LP2023.09 | 4.8 | 6.18 | 4.87 | 5.28 | — | — | — | — | |
| FSA-NetOwn implementation=true, Wider rotation range=false, Training dataset=300W-LP2023.09 | 4.83 | 6.25 | 4.94 | 5.34 | 6.88 | 6.52 | 7.28 | 6.89 | |
| WHENetOwn implementation=false, Wider rotation range=true, Training dataset=300W-LP2023.09 | 5.11 | 6.24 | 4.92 | 5.42 | — | — | — | — | |
| HopeNetOwn implementation=true, Wider rotation range=false, Training dataset=300W-LP2023.09 | 6.4 | 6.53 | 5.39 | 6.11 | 7.98 | 6.4 | 8.54 | 7.64 |