实验室硕士生王自强针对视频注视点预测提出基于时空自注意力机制的三维卷积神经网络(STSANET)近期在国际期刊 IEEE Transactions on Multimedia(TMM)上发表。

[1]Z. Wang et al., "Spatio-Temporal Self-Attention Network for Video Saliency Prediction," in IEEE Transactions on Multimedia, doi: 10.1109/TMM.2021.3139743.

Code: https://github.com/WZq975/STSANet