1 paper
Tian Xia, Zihan Ma, Xinlong Wang +4
Decoding images from fMRI often involves mapping brain activity to CLIP's final semantic layer. To capture finer visual details, many approaches add a parameter-intensive VAE-based…