基于对比学习的声源定位引导视听分割模型
|
|
黄文湖,赵邢,谢亮,梁浩然,梁荣华
|
Contrastive learning-based sound source localization-guided audio-visual segmentation model
|
|
Wenhu HUANG,Xing ZHAO,Liang XIE,Haoran LIANG,Ronghua LIANG
|
|
| 表 12 损失系数对模型性能的影响 |
| Tab.12 Impact of loss coefficients on model performance |
|
| 编号 | $ {\lambda _{\text{1}}} $ | $ {\lambda _{\text{2}}} $ | $ {\lambda _{\text{3}}} $ | MS3 | | $ {M_{\text{J}}} $/% | $ {M_{\text{F}}} $/% | | 0 | 0.00 | 0.00 | 0.00 | 56.67 | 67.4 | | 1 | 0.00 | 0.00 | 0.05 | 56.86 | 68.1 | | 2 | 0.00 | 0.00 | 0.10 | 57.62 | 68.1 | | 3 | 0.00 | 0.00 | 0.50 | 56.88 | 67.1 | | 4 | 0.00 | 0.05 | 0.10 | 59.07 | 69.5 | | 5 | 0.00 | 0.10 | 0.10 | 59.11 | 69.8 | | 6 | 0.00 | 0.50 | 0.10 | 57.72 | 68.2 | | 7 | 0.01 | 0.10 | 0.10 | 59.50 | 69.8 | | 8 | 0.05 | 0.10 | 0.10 | 55.43 | 66.6 | | 9 | 0.10 | 0.10 | 0.10 | 57.55 | 68.1 |
|
|
|