| Benchmark | Model | Rank | Results |
|---|---|---|---|
| audio-visual-active-speaker-detection-on-ava | SPELL+ | #4 | validation mean average precision: 94.9% |
| audio-visual-active-speaker-detection-on-ava | SPELL | #5 | validation mean average precision: 94.2% |
| node-classification-on-ava | ASDNet [ASDNet_ICCV2021] | #1 | mAP: 93.5 |
| node-classification-on-ava | TalkNet [tao2021someone] | #2 | mAP: 92.3 |
| node-classification-on-ava | UniCon [zhang2021unicon] | #3 | mAP: 92 |
| node-classification-on-ava | MAAS-TAN [MAAS2021] | #4 | mAP: 88.8 |