# Compute-Efficient Medical Image Classification with Softmax-Free Transformers and Sequence Normalization

arXiv: https://arxiv.org/abs/2406.01314 · HTML: https://mlsanity.com/paper/2406.01314/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 14 of its 17 references that are in the map.
- First posted on arXiv 3 Jun 2024.

## Builds on (most cited first, 14 of 14)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift](https://mlsanity.com/paper/1502.03167/) (2015)
- [Transformers are RNNs: Fast Autoregressive Transformers with Linear Attention](https://mlsanity.com/paper/2006.16236/) (2020)
- [Rethinking Attention with Performers](https://mlsanity.com/paper/2009.14794/) (2020)
- [Instance Normalization: The Missing Ingredient for Fast Stylization](https://mlsanity.com/paper/1607.08022/) (2016)
- [Nystr"omformer: A Nystr"om-Based Algorithm for Approximating Self-Attention](https://mlsanity.com/paper/2102.03902/) (2021)
- [Transformers in Medical Imaging: A Survey](https://mlsanity.com/paper/2201.09873/) (2022)
- [cosFormer: Rethinking Softmax in Attention](https://mlsanity.com/paper/2202.08791/) (2022)
- [A Patient-Centric Dataset of Images and Metadata for Identifying Melanomas Using Clinical Context](https://mlsanity.com/paper/2008.07360/) (2020)
- [SOFT: Softmax-free Transformer with Linear Complexity](https://mlsanity.com/paper/2110.11945/) (2021)
- [AIROGS: Artificial Intelligence for RObust Glaucoma Screening Challenge](https://mlsanity.com/paper/2302.01738/) (2023)
- [SimA: Simple Softmax-free Attention for Vision Transformers](https://mlsanity.com/paper/2206.08898/) (2022)
- [Cascaded Cross-Attention Networks for Data-Efficient Whole-Slide Image Classification Using Transformers](https://mlsanity.com/paper/2305.06963/) (2023)

## Abstract (arXiv)

The Transformer model has been pivotal in advancing fields such as natural language processing, speech recognition, and computer vision. However, a critical limitation of this model is its quadratic computational and memory complexity relative to the sequence length, which constrains its application to longer sequences. This is especially crucial in medical imaging where high-resolution images can reach gigapixel scale. Efforts to address this issue have predominantely focused on complex techniques, such as decomposing the softmax operation integral to the Transformer's architecture. This paper addresses this quadratic computational complexity of Transformer models and introduces a remarkably simple and effective method that circumvents this issue by eliminating the softmax function from the attention mechanism and adopting a sequence normalization technique for the key, query, and value tokens. Coupled with a reordering of matrix multiplications this approach reduces the memory- and compute complexity to a linear scale. We evaluate this approach across various medical imaging datasets comprising fundoscopic, dermascopic, radiologic and histologic imaging data. Our findings highlight that these models exhibit a comparable performance to traditional transformer models, while efficiently handling longer sequences.