Azimuth Classification on Pix3D (test)
77.93Bed AccuracyOurs
Evaluation results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Oursnumber of bins=9, overlap=2.5°, training protocol=category agnostic2022.03 | 77.93 | 74.55 | 59.09 | 61.39 | 85.06 | 90.74 | 83.54 | 30 | 36.36 | 73.56 | 76.08 | |
| Ours_phase1number of bins=9, overlap=2.5°, training protocol=category agnostic2022.03 | 75.12 | 70.61 | 62.34 | 60.43 | 88.19 | 94.44 | 84.81 | 40 | 27.27 | 72.21 | 76.08 | |
| Mousavian et al.number of bins=9, overlap=2.5°, training protocol=category agnostic2022.03 | 71.83 | 47.99 | 51.3 | 56.59 | 80.96 | 75.93 | 68.35 | 15 | 18.18 | 57.87 | 75.81 | |
| ResNet_baselinenumber of bins=9, overlap=2.5°, training protocol=category agnostic2022.03 | 68.08 | 58.95 | 50.65 | 51.8 | 81.2 | 72.22 | 67.09 | 20 | 27.27 | 61.78 | 75.25 | |
| Baselinenumber of bins=9, overlap=2.5°, training protocol=category agnostic2022.03 | 65.73 | 50.73 | 41.56 | 51.08 | 78.55 | 77.78 | 65.82 | 20 | 18.18 | 56.72 | 69.45 |