Jongchan/attention-module
PyTorch implementation of two attention mechanisms (BAM, CBAM) for improving CNN performance on visual recognition tasks.
Not currently ranked — collecting fresh signals.
star history
This repository provides official PyTorch implementations of BAM (Bottleneck Attention Module) and CBAM (Convolutional Block Attention Module), two attention mechanisms designed to improve convolutional neural networks. Both modules can be plugged into existing CNN architectures like ResNet to boost performance on image classification. The code includes training scripts for ImageNet, pretrained checkpoints, and can be extended to other vision tasks.
Frequently asked
- What is Jongchan/attention-module?
- PyTorch implementation of two attention mechanisms (BAM, CBAM) for improving CNN performance on visual recognition tasks.
- Is attention-module open source?
- Yes — Jongchan/attention-module is open source, released under the MIT license.
- What language is attention-module written in?
- Jongchan/attention-module is primarily written in Python.
- How popular is attention-module?
- Jongchan/attention-module has 2.2k stars on GitHub.
- Where can I find attention-module?
- Jongchan/attention-module is on GitHub at https://github.com/Jongchan/attention-module.