Indoor/Outdoor Classification of Voice for Mobile Devices

This project’s goal is to classify a two-way radio recording into one of two classes: “indoor” or “outdoor” recording.
A literature review that we have made led us to choose a neural network that was designed for solving a similar problem. The network is a ResNet-based network. The system transforms audio signals to log-mel spectrograms, and the result is then classified by the network.
Since the data base we have is too small, another goal was defined for the project: training on a bigger dataset and perform inference on the small one.
In order to achieve this goal, we wrote a few preprocessing functions that make the data bases more similar in their features, and we used transfer learning to complete the train.
Unfortunately, our results show that this method did not achieve the goal.
Finally, the optimal results were obtained by using the chosen network on the small data set. We have reached 96.3% success rate.

Indoor/Outdoor Classification of Voice for Mobile Devices

Collaboration:

Refael logo