Adding RNN-LSTM architecture

Feature requests
#1

hi.
are there any plans to add the RNN-LSTMmodel in the near future?
I thought it would be useful to add an option for audio classification.

regards,

#2

Hi @desnug1 ,

The enterprise feature enables organisations and developers to add their own custom learning blocks from any framework. This means you can implement the RNN-LSTM model in the edge impulse studio using this feature. See here for more info Adding custom learning blocks

Regards,
Clinton

#3

hi @oduor_c
thanks for the reply. I didn’t know that feature existed in the enterprise.

  1. how much does it cost? is there a special price for students?
  2. based on your experience, which one is perform better between CNN or RNN for classifying audio?
#4

Hi @desnug1 ,

For enquiries about pricing and whether there is a special price for students, kindly contact customer support through hello@edgeimpulse.com . Regarding CNN vs RNN for audio classification, CNNs have worked perfectly for me and I haven’t yet personally tried RNNs so I can’t currently suggest which is better in that case.

Regards,
Clinton