YangGangZhiQi

谢谢作者在图像风格转换任务上的杰出工作。介绍里面的“officially unofficial”应该怎么翻译呢？官方的非正式？

Hi, thanks for your great works. I wonder whether the CRD distillation loss could be used in image enhancement tasks like Denoise\SR\Deblur. I would appreciate it if someone could answer...

Hi, thanks for this great work. After reading codes, I have a question about self.mask_token in SimMIM projects. What the role of self.mask_token in the init func of SwinTransformerForSimMIM? It...

Hi, thanks for this great work. After reading codes, I have a question about self.mask_token in SimMIM projects. What the role of self.mask_token in the init func of SwinTransformerForSimMIM? It...

Hi, thanks for this great work. When training pvtv2 detection model with config retinanet_pvt_v2_b2_fpn_1x_coco.py. I got some strange information from mmcv，and do not know whether it is harmful for training...

Hi, thanks for this great works. After reading the source code, I have a question about TV_3D module. Why do we need to multiply 2.0 in self.weight_r\self.weight_g\self.weight_b ? What's the...

Hi, thanks for this great works! When reading the source code of CoOp，I have an question about 'token_prefix' and 'token_suffix' in PromptLearner class. How to under this two variables? ...

Thanks for the awesome implementation! After training in my own datasets, the classification acc1 of model is bad. I do not know how to analysis the problem. When finetuning CLIP,...

Selecting query-key patches

2

@lyndonzheng Hi, thanks for your great work. After reading your paper and source code, I have a question about selecting query-key patches. In your source codes, the selecting codes are...

Thanks for your great works! After reading the papers and codes，I have a question about Contrastive loss。In your codes, using the unrelated samples in a batch as negative samples. How...