Unsupervised Learning of Semantic Audio Representations

Abstract

Even in the absence of any explicit semantic annotation, vast collections of audio recordings provide valuable information for learning the categorical structure of sounds. We consider several class-agnostic semantic constraints that apply to unlabeled nonspeech audio: (i) noise and translations in time do not change the underlying sound category, (ii) a mixture of two sound events inherits the categories of the constituents, and (iii) the categories of events in close temporal proximity are likely to be the same or related. Without labels to ground them, these constraints are incompatible with classification loss functions. However, they may still be leveraged to identify geometric inequalities needed for triplet loss-based training of convolutional neural networks. The result is low-dimensional embeddings of the input spectrograms that recover 41% and 84% of the performance of their fully-supervised counterparts when applied to downstream query-by-example sound retrieval and sound event classification tasks, respectively. Moreover, in limited-supervision settings, our unsupervised embeddings double the state-of-the-art classification performance.

Submitted to ICASSP 2018

References in corpus (6)

• Convolutional Recurrent Neural Networks for Polyphonic Sound Event Detection
• SoundNet: Learning Sound Representations from Unlabeled Video
• Unsupervised Learning of Visual Representations using Videos
• Learning to See by Moving
• Deep Convolutional Neural Networks and Data Augmentation for Acoustic Event Detection
• Look, Listen and Learn

Cited by in corpus (4)

• Towards Learning a Universal Non-Semantic Representation of Speech
• Gradient-based Training of Slow Feature Analysis by Differentiable Approximate Whitening
• Improved Audio Embeddings by Adjacency-Based Clustering with Applications in Spoken Term Detection
• Phonetic-and-Semantic Embedding of Spoken Words with Applications in Spoken Content Retrieval