Two Papers at ICASSP 2021

We have two new papers at ICASSP 2021 this year:

• “Few-Shot Continual Learning for Audio Classification” led by Yu Wang. This paper investigates how to expand the vocabulary of an existing audio tagging model when you only have a few examples of the new class. It will be presented at ICASSP on June 9, 3:30 - 4:15PM (EDT).
• “Specialized Embedding Approximation for Edge Intelligence: A Case Study in Urban Sound Classification” led by Sangeeta Srivastava. This paper investigates compressing an audio embedding network using knowledge distillation to a specialized target domain. It will be presented at ICASSP on June 9, 2:00 - 2:45PM (EDT).

Please check out these presentations and papers by Sangeeta and Yu!