CTAL: Pre-training Cross-modal Transformer for Audio-and-Language Representations

Abstract

Existing audio-language task-specific predictive approaches focus on building complicated late-fusion mechanisms. However, these models are facing challenges of overfitting with limited labels and low model generalization abilities. In this paper, we present a Cross-modal Transformer for Audio-and-Language, i.e., CTAL, which aims to learn the intra-modality and inter-modality connections between audio and language through two proxy tasks on a large amount of audio-and-language pairs: masked language modeling and masked cross-modal acoustic modeling. After fine-tuning our pre-trained model on multiple downstream audio-and-language tasks, we observe significant improvements across various tasks, such as, emotion classification, sentiment analysis, and speaker verification. On this basis, we further propose a specially-designed fusion mechanism that can be used in fine-tuning phase, which allows our pre-trained model to achieve better performance. Lastly, we demonstrate detailed ablation studies to prove that both our novel cross-modality fusion component and audio-language pre-training methods significantly contribute to the promising results.

arXiv commentThe 2021 Conference on Empirical Methods in Natural Language Processing

References in corpus (7)

• Efficient Estimation of Word Representations in Vector Space
• Distributed Representations of Sentences and Documents
• VisualBERT: A Simple and Performant Baseline for Vision and Language
• Mockingjay: Unsupervised Speech Representation Learning with Deep Bidirectional Transformer Encoders
• TERA: Self-Supervised Learning of Transformer Encoder Representation for Speech
• Speech Emotion Recognition Using Multi-hop Attention Mechanism
• Improving Transformer-based Speech Recognition Using Unsupervised Pre-training