VAD Regression on VOICES (test)
0.105MAEEye-Gaze Temporal CNN
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Eye-Gaze Temporal CNNInput duration=2 seconds2026.01 | 0.105 | 0.166 | |
| Eye-Gaze Temporal CNNInput duration=5 seconds2026.01 | 0.106 | 0.157 | |
| Eye-Gaze Temporal CNNInput duration=10 seconds2026.01 | 0.108 | 0.148 | |
| Statistical Features (eyes only)Input duration=2 seconds2026.01 | 0.11 | 0.157 | |
| Statistical Features (eyes only)Input duration=5 seconds2026.01 | 0.11 | 0.226 | |
| Statistical Features (eyes only)Input duration=10 seconds2026.01 | 0.11 | 0.223 | |
| Statistical Features (eyes + face)Input duration=2 seconds2026.01 | 0.11 | 0.167 | |
| Statistical Features (eyes + face)Input duration=5 seconds2026.01 | 0.11 | 0.228 | |
| Statistical Features (eyes + face)Input duration=10 seconds2026.01 | 0.11 | 0.225 | |
| GLASSInput duration=2 seconds, Pre-training forecasting window=10 seconds2026.01 | 0.119 | 0.285 | |
| GLASSInput duration=5 seconds, Pre-training forecasting window=5 seconds2026.01 | 0.122 | 0.284 | |
| GLASSInput duration=5 seconds, Pre-training forecasting window=10 seconds2026.01 | 0.122 | 0.294 | |
| GLASSInput duration=10 seconds, Pre-training forecasting window=10 seconds2026.01 | 0.122 | 0.297 | |
| GLASSInput duration=2 seconds, Pre-training forecasting window=5 seconds2026.01 | 0.123 | 0.283 | |
| GLASSInput duration=10 seconds, Pre-training forecasting window=5 seconds2026.01 | 0.125 | 0.285 | |
| GLASSInput duration=2 seconds, Pre-training forecasting window=2 seconds2026.01 | 0.152 | 0.23 | |
| GLASSInput duration=10 seconds, Pre-training forecasting window=2 seconds2026.01 | 0.154 | 0.167 | |
| GLASSInput duration=5 seconds, Pre-training forecasting window=2 seconds2026.01 | 0.161 | 0.228 |