->Developed a prediction tool of Notes and Instruments in a given Audio file in a team of four members.
->Used SVM classifier to detect the instrument being played in an audio file, along with all the notes and frequencies present in the audio file.
->Extended it to make a robotic system that would mimic the audio, based on the results given by the classifier.
->Reached in the pre-final round after clearing five competitive stages.
->Tools - Sklearn, NumPy, Pandas, Wave, Struct, ML Algorithms
Tasks accomplished:
Python:
Take music file as input.
Machine Learning to detect the instruments used.
Use audio processing to detect note onsets and individual notes.
Use serial communication to send this information to AtMega 2560.
Embedded C:
Bot:
Receive serial input from the Python script.
Process input and control the bot to strike on piano and trumpet.
Instrument:
Capacitive touch sensor(placed on piano and trumpet) receives input on striking.
Detect which note is being played.