Jian Xiao

icon indicating an email [email protected]

icon company Hefei University of Technology icon location Hefei, China icon location Computer Vision, Video Understanding, Multimodal Learning. MS student, Computer Science and Information Engineering， Hefei University of Technology

Results 4 issues of Jian Xiao

About LF-VILA code in PatchEmbed3D of video encoder

the padding seems not right, or maybe i made a mistake ``` # padding _, _, D, H, W = x.size() if H % self.patch_size[0] != 0: x = F.pad(x,...

about clipvip-vit-16 pretrained weights file

when finetuning on clipvip-vit-16 pretrained weights, it trains the model from recall@1 = 0, which means the provided pretrained weights file seems wrong. can you provide the correct clipvip-vit-16 pretrained...

About activitynet captions dataset in CLIP-ViP

hello, thank you for sharing your excellent work! I have reproduced result in msrvtt and even acquire a higher result than that in paper. But when I tried to reproduce...

question about QB-norm inference

Hello, I found that you used QB-Norm postprocessing in inference stage while there is no mention about qb-norm in paper, can you show the result without qb-norm? thank you for...