glazhh

I encountered a similar issue. I resolved it by adding `torch.cuda.synchronize()` before using encode_jpeg. It seems there might be some synchronization problems between `F.interpolate` and `torchvision.io.encode_jpeg`. ```python resized_image_tensor = F.interpolate(image_data.unsqueeze(0),...

@nagadomi to_uint8 still writing encode_jpeg input data while encoding it, consider change your code to this before this issue fixed ```Python frame_uint8 = to_uint8(frame) torch.cuda.synchronize() jpeg_data = encode_jpeg(frame_uint8, quality=quality) torch.cuda.synchronize()...