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 9 双向注意力融合模块对模型性能的影响

Tab.9 Impact of bidirectional attention fusion module on model performance

融合方式 S4 MS3 $ {M_{\text{J}}} $/% $ {M_{\text{F}}} $/% $ {M_{\text{J}}} $/% $ {M_{\text{F}}} $/% 不进行任何融合 77.29 86.9 55.98 66.9 仅视觉融合 77.66 87.3 58.84 69.0 仅音频融合 77.84 87.2 57.57 68.6 双向融合 78.68 88.0 59.50 69.8