Closed yinxu1996 closed 3 months ago
The input is visual neural activities
---- Replied Message ---- | From | Yin @.> | | Date | 07/20/2024 11:04 | | To | @.> | | Cc | @.***> | | Subject | [hw654833612/Visual-Language-Decoding] training and testing (Issue #1) |
Dear author, hello! I have a question that troubles me: your tasks are catagory decoding and labeling decoding, but you use image category and labeling as inputs to train the model? What are the inputs for the model testing phase?
— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you are subscribed to this thread.Message ID: @.***>
Dear author, hello! I have a question that troubles me: your tasks are catagory decoding and labeling decoding, but you use image category and labeling as inputs to train the model? What are the inputs for the model testing phase?