HearingTechnology / CoNNear_IHC-ANF

A CNN model of the human IHC and ANF complex
Other
14 stars 4 forks source link

CNN models #1

Closed wzhCUMT closed 2 years ago

wzhCUMT commented 2 years ago

Dear Mr./Ms. Fotios Drakopoulos and Relevant researchers ,

@fotisdr @HearingTechnology Thank you very much for shared the excellent code! Could you please share the CNN code of CoNNear(cochlea), CoNNear(IHC–ANF), and CoNNear(IHC–ANF). Sincerely thank you for your help!

Best Wishes, wzh

fotisdr commented 2 years ago

Dear wzh,

What do you mean by the CNN code? The CNN model of CoNNearcochlea can be found in the connear folder of the CoNNear_cochlea repository, while the CNN models of the CoNNearIHC and CoNNearANF can be found in the connear folder of this repository (json and h5 files that can be loaded in Tensorflow/Keras). Example scripts can also be found in both repositories on how to run the models.

Best, Fotios

wzhCUMT commented 2 years ago

Dear Mr. Fotios,

Thanks a lot for your quick reply! I want to learn how you use tensorflow to build the CNN code of CoNNear_cochlea, CoNNearIHC, and CoNNearANF. However, I can only see the structure and parameters of the CNN model by loading the josn and h5 files in the connear folder. So, could you please share it.

Cheers, wzh

fotisdr commented 2 years ago

Dear wzh,

I see, sorry for the misunderstanding. Unfortunately the Python scripts that were used to build and train each CoNNear model are not available at the moment, but the exact implementations of the CoNNear models are described in the published CoNNear_cochlea and CoNNear_IHC-ANF papers. You can contact me if you need additional help with that. Additionally, we plan on making public a repository that contains the whole CoNNear periphery model (cochlea+IHC+ANF) on the HearingTechnology profile the following days, perhaps this might be of help for you.

Cheers, Fotios

EDIT: By the way, you can find the aforementioned repository here: https://github.com/HearingTechnology/CoNNear_periphery

wzhCUMT commented 2 years ago

Very nice! Thank you very much for your help!