Microsoft & Bath U’s SpectFormer Significantly Improves Vision Transformers via Frequency and Attention

In the new paper SpectFormer: Frequency and Attention Is What You Need in a Vision Transformer, a research team from Microsoft and the University of Bath proposes Spectformer, a novel transformer architecture that combines spectral and multi-headed attention layers to better capture appropriate feature representations and improve performance.

ai architecture artificial intelligence attention attention mechanisms computer vision & graphics deep-neural-networks feature machine learning machine learning & data science microsoft ml novel paper performance research research team team technology transformer transformer architecture transformers university vision vision-transformer