Environment Transfer for Distributed Systems
This paper tackles the problem of data scarcity for distributed acoustic machine learning, which affects the robustness and generalizability of models in varied acoustic environments.
The paper addresses the problem of insufficient data for distributed acoustic machine learning by proposing a method to transfer acoustic style textures between environments. This method generates augmented audio data that exhibits transferred environmental features while preserving content features, as evaluated by classification accuracy and content preservation metrics.
Collecting sufficient amount of data that can represent various acoustic environmental attributes is a critical problem for distributed acoustic machine learning. Several audio data augmentation techniques have been introduced to address this problem but they tend to remain in simple manipulation of existing data and are insufficient to cover the variability of the environments. We propose a method to extend a technique that has been used for transferring acoustic style textures between audio data. The method transfers audio signatures between environments for distributed acoustic data augmentation. This paper devises metrics to evaluate the generated acoustic data, based on classification accuracy and content preservation. A series of experiments were conducted using UrbanSound8K dataset and the results show that the proposed method generates better audio data with transferred environmental features while preserving content features.