Hello, unfortunately this code is not maintained anymore and Theano itself has been discountinued at the end of 2017. I am not sure where you could find the pretrained weights for VGG for Theano now, but it would probably be best to reproduce ReSeg in an actively developed framework such as JAX, Tensorflow or PyTorch rather than trying to get this code to work as is.
Hello, unfortunately this code is not maintained anymore and Theano itself has been discountinued at the end of 2017. I am not sure where you could find the pretrained weights for VGG for Theano now, but it would probably be best to reproduce ReSeg in an actively developed framework such as JAX, Tensorflow or PyTorch rather than trying to get this code to work as is.