Training model
The trained model can be public? Is it consistent with the results in the paper?
The sufflenetv2 model trianed on ic13 and15 has been released, but for some implementation details are different, the results are not as good as the results in the paper. I will continue to update this project with the reference to the official implementation.
Does this only work on GPU?
It can work on CPU, by setting the gpu_id=None and uncommenting torch.cuda.synchronize(self.device).
This result looks not as good as in the paper
yes, i am working hard to improve the performance. I will release a better version some days later.
@SURFZJY Did the results got better? whats the result on icdar15