Low-complexity deep learning frameworks for acoustic scene classification using teacher-student scheme and multiple spectrograms
This work addresses the problem of efficient acoustic scene classification for resource-constrained applications, but it is incremental as it builds on existing teacher-student distillation methods.
The paper tackles acoustic scene classification by proposing a low-complexity deep learning system using a teacher-student scheme and multiple spectrograms, achieving a classification accuracy of 57.4% and improving the DCASE baseline by 14.5%.
In this technical report, a low-complexity deep learning system for acoustic scene classification (ASC) is presented. The proposed system comprises two main phases: (Phase I) Training a teacher network; and (Phase II) training a student network using distilled knowledge from the teacher. In the first phase, the teacher, which presents a large footprint model, is trained. After training the teacher, the embeddings, which are the feature map of the second last layer of the teacher, are extracted. In the second phase, the student network, which presents a low complexity model, is trained with the embeddings extracted from the teacher. Our experiments conducted on DCASE 2023 Task 1 Development dataset have fulfilled the requirement of low-complexity and achieved the best classification accuracy of 57.4%, improving DCASE baseline by 14.5%.