The aim of the project is to create an Active Deep Learning and Machine Learning model that would predict the class of the calls depending on the model the user selects. It would perform active learning while training by querying the user to annotate the data that the model has identified with lower confidence(other querying strategies would also be used). This data annotated by the user is again used for retraining the model. The prediction of any class by the model that the user desires would be based on the spectrograms generated by the audio files.



Kunal Mehta


  • Jesse Lopez
  • Scott Veirs
  • Valentina
  • Abhishek_Singh
  • val veirs