haoyi-duan / dg-sct Goto Github PK

Hi, do you plan to make your paper public in arxiv so that others can cite it?

Hi, could I know when you will release your paper? Many thanks

Hi, could you share the number of trainable and total parameters of your model on AVE, AVS, and AVQA tasks? Many thanks.

Hi,

Thanks for the work! I'm trying to download the pre-trained checkpoints for the backbone model, using the google drive link, but it requires permission. Is it possible to change the access type of the file, since I guess many others would like to get them as well?

Best,
Arthur

Hi, could I know how you get the pre-trained weights of grounding module? I found you use the different pre-trained weights from LAVISH. Do you use another dataset to pre-trained the whole model to get the pre-trained ground module? If yes, which dataset you use for this pre-training? Many thanks.

Hi, I am wondering which parts of parameters in AVQA are trainable and forzen? Whether only the parameters of pre-trained ViTs are frozen and others including your proposed prompts, ground modules, and text encoder are trainable? Many thanks.