Methods › General › Attention Modules › Re-Attention Module
Re-Attention Module
archive 2025-07-28 Description, source and code snippet are the archive's method entry.
The Re-Attention Module is an attention layer used in the DeepViT architecture which mixes the attention map with a learnable matrix before multiplying with the values. The motivation is to re-generate the attention maps to increase their diversity at different layers with negligible computation and memory cost. The authors note that traditional self-attention fails to learn effective concepts for representation learning in deeper layers of ViT -- attention maps become more similar and less diverse in deeper layers (attention collapse) - and this hinders the model from getting expected performance gain. Re-attention is implemented by:
Re-Attention(Q, K, V)=Norm(Θ^⊤(Softmax((Q K^⊤)/(√(d))))) V
where transformation matrix Θ is multiplied to the self-attention map A along the head dimension.
Papers archive 2025-07-28
4 shown of 4, newest first. Repository counts are the archive's code-links table. A Syntology line states what Syntology ran from that paper's harvested code; it is per sample and not a correctness claim.
-
ODC-SA Net: Orthogonal Direction Enhancement and Scale Aware Network for Polyp Segmentation 10 May 2024 · 0 repositories · arXiv:2405.06191
-
Learning to Agree on Vision Attention for Visual Commonsense Reasoning 4 Feb 2023 · 0 repositories · arXiv:2302.02117
-
SAT: Size-Aware Transformer for 3D Point Cloud Semantic Segmentation 17 Jan 2023 · 0 repositories · arXiv:2301.06869
-
DeepViT: Towards Deeper Vision Transformer 22 Mar 2021 · 5 repositories · arXiv:2103.11886Syntology ran 1 of 1 samples · 0 unverified
Tasks archive 2025-07-28
8 tasks the archive attaches to papers tagged with this method, by distinct papers. A task without a page in the catalog is plain text.
| Task | Papers |
|---|---|
| Image Classification | 1 |
| Point Cloud Segmentation | 1 |
| Representation Learning | 1 |
| Segmentation | 1 |
| Semantic Segmentation | 1 |
| Visual Commonsense Reasoning | 1 |
| Visual Reasoning | 1 |
| image-classification | 1 |
Usage over time archive 2025-07-28
Components: the archive holds no method-to-method composition, so PwC's Components table cannot be rebuilt; the Papers list carries no Results column for the same reason (the archive does not join its leaderboard rows to method tags).
Categories archive 2025-07-28
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections