search

layumi / 2016_super_resolution

ICCV2015 Image Super-Resolution Using Deep Convolutional Networks
MIT License
86 stars 43 forks source link

I have one more question about "train objective" and "val objective" in result after training by running "train_SRnet.m" #13

Open star4s opened 7 years ago

star4s commented 7 years ago

I have one more question about "train objective" and "val objective" in result after training by running "train_SRnet.m"

After I take training by running "train_SRnet.m", I got a bottom graph image.

1

What meaning is the gap between "train objective" and "val objective"? The "train objective" blue line is bottom against the "val objective" red line over epoch 3. I am so sorry every time, I am bother you. Thank you very much for your attention.

layumi commented 7 years ago
  1. Train: CNN use the training data to tune the parameters so it usually over-fits the training set. (show a lower loss value)
  2. Validation: CNN did not see the validation data before. The validation result can reflect the "real" situation in the test setting. So it usually achieves a little bit higher loss.