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 3 去除池化层引起的特征尺寸变化对实验结果的影响

Tab.3 Effect of characteristic size change caused by removal of pool layer on experimental results

去除/保留 最大池 化层 AVSegFormer SSL2AVS S4 MS3 S4 MS3 $ {M_{\text{J}}} $/% $ {M_{\text{F}}} $/% $ {M_{\text{J}}} $/% $ {M_{\text{F}}} $/% $ {M_{\text{J}}} $/% $ {M_{\text{F}}} $/% $ {M_{\text{J}}} $/% $ {M_{\text{F}}} $/% 去除 76.11 86.4 43.41 58.0 77.16 86.8 56.18 66.9 保留 76.45 85.9 49.53 62.8 76.87 86.5 52.49 63.3