TCSVT-2022-BVQA
TCSVT-2022-BVQA copied to clipboard
zwx8981
Blindly Assess Quality of In-the-Wild Videos via Quality-aware Pre-training and Motion Perception
SpatialExtractor/get_spatialextractor_model.py#L9 model = torch.nn.DataParallel(model).cuda() 这个模型配置了多GPU并行，为什么我有2个GPU，但运行时还是只在cuda 0上？
Hello. Thanks for your great work. I am trying to retrain the model with LSVQ datasets. However, the form of LSVQ is disabled, and I can not get the datasets....
Hi, I'm trying to use this VQA architecture to predict a target other than MOS. My target label ranges from 10 to 41. How should I change the VQAModel's forward...
Hi, You have used `.cuda()` in many places in your code without checking if the device has any GPUs or not. Therefore, your code does not run on CPU. So...
How should I change the code, if I want to use frame_batch_size=1 ? If I just change it to 1, The row size multiplied by 32 is still output..
