baccuslab / deep-retina Goto Github PK

hello, like your repository. But can you provide some images dataset and pre-trained models?

thank you very much~

Hi, we are now doing further research based on your excellent work, could you please provide the training data example or the pretrained model ? Thanks~

Apparently super() needs at least one argument. I tried just using "self" but it "needs a type, not convnet".

At line 195 in deepretina/models.py .format() receives three different variables (iteration, loss, elapsed_time), but when printing the string iteration is just copied three times. Need to revisit in the morning.

Hello,

In experiments.py (line 78/80) you normalize the experiment data immediately upon loading it. Shouldn't you first split the data into train and validation sets before calculating normalization statistics?

Best,
Daniel

Hello,

I do like your repository. Can you provide some pre-trained models or some data sets for training?

Thank you so much in advance.

Best wishes,
Guangyao

Hi,

Is it possible to have access to the trained model or the dataset you used to train the model?

Thanks

I'm trying to run your runme.py script, but it breaks on importing convnet and sequential from deepretina.model.

Should it be nips_cnn instead of convnet, or is that something different that has since been removed?
Also, are you referring to Keras' standard Sequential instead of sequential, or did you make some adjustments there?

Thanks,
Michael

Likely just a mismatch in input/output sizes, but it's annoying to track down

Hello,

Is there an example to be able to use your neural network with natural images? For example, by using some images or video files as the input. I'm thinking in a way to use your fantastic tool in the same manner that the Virtual-Retina framework (https://link.springer.com/article/10.1007/s10827-008-0108-4). I guess you know what I'm talking about :)

Thank you so much in advance.

Kind regards,
Rubén