Hiusam

[Question] Usage with Multimodal LLM

12

Dear Authors, Thank you so much for your wonderful work. I want to ask if I am running LLaVA(https://github.com/haotian-liu/LLaVA/blob/main/llava/model/llava.py), a multimodal LLM built upon LLaMA by adding an image encoder,...

Why do you have a different results of object classifier for Nr3D and Sr3D

2

Hi, in your paper: > We note that in Nr3D (Sr3D) test accuracies for the two classifiers are 93.0% (100.0%) and 64.7% (67.4%), indicating a noticeable asymmetry in the difficulty...

Hi Junbo, In your forward_ssl code, you squeeze the fps sampling sequence, but this will raise errors when the batch size is 1 because you index the fps_choice with the...

wait for code

7

Hey! The idea in your paper is attractive! Waiting for your code now. :)