Inspiration
Building an audio classifier using CNNs and distinguishing between animal sounds with a high accuracy.
What it does
Distinguishes between the sounds of a dog and cat.
How we built it
Using librosa, numpy, keras, tensorflow and implementing CNNs followed by deployment using a .hdf5 file for weights.
Challenges we ran into
We were having some issues using the .h5 model and in pickling, but we overcame that using a .hdf5 file for deployment.
Accomplishments that we're proud of
We ended up getting a 100% training accuracy and a 91.07% testing accuracy.
What we learned
We learned how to work with flask and deploy the model and using librosa to create the mfcc series and getting an image-like data from that.
What's next for Dog Cat Sound Classifier
A classifier for dog cat counds
Log in or sign up for Devpost to join the conversation.