ChAda-ViT : Channel Adaptive Attention for Joint Representation Learning of Heterogeneous Microscopy Images

New Preprint out ๐ŸŽ‰

Introduction of ChAda-ViT a new Channel Adaptive Vision Transformer architecture integrating an Inter-Channel Attention mechanism that effectively handles images with varying numbers, orders, and types of channels and specially designed for biological images. A first step towards foundational model in biological imaging.

https://arxiv.org/abs/2311.15264

3