Search Results for author:
Found 3 papers, 1 papers with code
Learnable Irrelevant Modality Dropout for Multimodal Action Recognition on Modality-Specific Annotated Videos
However, effectively leveraging the audio modality in vision-specific annotated videos for action recognition is of particular challenge.
EPSANet: An Efficient Pyramid Squeeze Attention Block on Convolutional Neural Network
Recently, it has been demonstrated that the performance of a deep convolutional neural network can be effectively improved by embedding an attention module into it.
Expansions in multiple bases over general alphabets
Expansions in non-integer bases have been extensively investigated since a pioneering work of R\'enyi.
Combinatorics Number Theory 11A63 (Primary) 37B10 (Secondary)
