landskape-ai/triplet-attention
PyTorch implementation of triplet attention, a novel cross-dimensional attention module for convolutional neural networks.
Not currently ranked — collecting fresh signals.
star history
This repository provides the official implementation of the Triplet Attention module from the WACV 2021 paper. The method uses a three-branch structure with rotation operations to capture inter-dimensional dependencies in feature tensors, improving channel-spatial interaction without significant computational overhead. The module is designed as a drop-in enhancement for CNNs and evaluates on ImageNet classification and object detection tasks.
Frequently asked
- What is landskape-ai/triplet-attention?
- PyTorch implementation of triplet attention, a novel cross-dimensional attention module for convolutional neural networks.
- Is triplet-attention open source?
- Yes — landskape-ai/triplet-attention is open source, released under the MIT license.
- What language is triplet-attention written in?
- landskape-ai/triplet-attention is primarily written in Jupyter Notebook.
- How popular is triplet-attention?
- landskape-ai/triplet-attention has 442 stars on GitHub.
- Where can I find triplet-attention?
- landskape-ai/triplet-attention is on GitHub at https://github.com/landskape-ai/triplet-attention.