Towards Learning Universal Audio Representations
This work addresses the need for general sound understanding in applications, but it is incremental as it builds on existing architectures and objectives.
The authors tackled the problem of learning universal audio representations for diverse tasks by introducing a benchmark and evaluating existing models, finding that a novel normalizer-free Slowfast NFNet achieved state-of-the-art performance across all domains.
The ability to learn universal audio representations that can solve diverse speech, music, and environment tasks can spur many applications that require general sound content understanding. In this work, we introduce a holistic audio representation evaluation suite (HARES) spanning 12 downstream tasks across audio domains and provide a thorough empirical study of recent sound representation learning systems on that benchmark. We discover that previous sound event classification or speech models do not generalize outside of their domains. We observe that more robust audio representations can be learned with the SimCLR objective; however, the model's transferability depends heavily on the model architecture. We find the Slowfast architecture is good at learning rich representations required by different domains, but its performance is affected by the normalization scheme. Based on these findings, we propose a novel normalizer-free Slowfast NFNet and achieve state-of-the-art performance across all domains.