Project-ListenBot (Sound Source Classification and Localization)

This repo contains code for the ListenBot project [Code to record, classify, localize sounds and then move Misty II robot towards the sound source.]

To run the code, it is preferable to first create an Anaconda environment and install various packages. The minimum packages to be isntalled are:

1. Packages related to Misty II Python SDK: https://pypi.org/project/Misty-SDK/. The detailed instructions and examples can be found at these links: https://prabhjotkaurgosal.com/getting-started-with-misty-ii-robot-and-its-python-sdk-page2/ and https://github.com/MistyCommunity/Python-SDK
2. Packages related to sound classification using PANNs Inference. The instructions can be found here: https://github.com/qiuqiangkong/panns_inference

After the set up is complete, the code can be run from the main file. As an example, the conda environment created is called SoundClassification_Misty.

1. Navigate to the directroy where the main file and helper classes are present using the terminal

• $cd Desktop
• $cd SoundClassification

2. Activate the conda environment

• $conda activate SoundClassification_Misty

Then, launch the Visual Studio Code.

References:

1. https://pypi.org/project/Misty-SDK/ - Link to Python SDK for Misty II
2. https://github.com/MistyCommunity/Python-SDK - Examples to run Python SDK for Misty II
3. https://prabhjotkaurgosal.com/getting-started-with-misty-ii-robot-and-its-python-sdk-page2/ - More detailed steps on installation and getting Misty II up and running using the Python SDK
4. https://github.com/qiuqiangkong/panns_inference - PANNs inference for sound classification