Recently, I have been using your metaRec model, and I want to add a multi-modal feature fusion mechanism to it. The initial idea is to embed the multi-modal features that have been processed by fusion into the original item features, I wonder how to improve it
Recently, I have been using your metaRec model, and I want to add a multi-modal feature fusion mechanism to it. The initial idea is to embed the multi-modal features that have been processed by fusion into the original item features, I wonder how to improve it