hulianyuyy / CorrNet

Continuous Sign Language Recognition with Correlation Network (CVPR 2023)
84 stars 14 forks source link

Ask #50

Open Anysc-Bryant opened 1 month ago

Anysc-Bryant commented 1 month ago

Hello, I encountered some confusion while using netron to inspect the model structure. Could you please clarify if the input dimension of 1024x1296 for the provided pre-trained model pertains to the image size? Additionally, could you explain what is meant by 'output' in this context

hulianyuyy commented 1 month ago

Sorry, but i don't quite understand where is the 1024x1296 come from. Our model uses 224*244 images as inputs to process, as noted in the dataloader_video.py.