1 code implementation • 9 Oct 2023 • Xianming Gu, Lihui Wang, Zeyu Deng, Ying Cao, Xingyu Huang, Yue-Min Zhu
Specifically, we propose the cross-attention fusion (CAF) block, which adaptively fuses features of two modalities in the spatial and frequency domains by exchanging key and query values, and then calculates the cross-attention scores between the spatial and frequency features to further guide the spatial-frequential information fusion.
no code implementations • 25 Aug 2023 • Yuxiao Luo, Ziyu Lyu, Xingyu Huang
In this paper, we propose a Time-Frequency Enhanced Decomposed Network (TFDNet) to capture both the long-term underlying patterns and temporal periodicity from the time-frequency domain.
no code implementations • CVPR 2020 • Dong Cao, Xiangyu Zhu, Xingyu Huang, Jianzhu Guo, Zhen Lei
Finally, we propose a Domain Balancing Margin (DBM) in the loss function to further optimize the feature space of the tail domains to improve generalization.