Integrating Self-Supervised Speech Model with Pseudo Word-Level Targets from Visually-Grounded Speech Model
Citations Over Time
Abstract
Recent advances in self-supervised speech models have shown significant improvement in many downstream tasks. However, these models predominantly centered on frame-level training objectives, which can fall short in spoken language understanding tasks that require semantic comprehension. Existing works often rely on additional speech-text data as intermediate targets, which is costly in the real-world setting. To address this challenge, we propose Pseudo-Word HuBERT (PW-HuBERT), a framework that integrates pseudo word-level targets into the training process, where the targets are derived from a visually-ground speech model, notably eliminating the need for speech-text paired data. Our experimental results on four spoken language understanding (SLU) benchmarks suggest the superiority of our model in capturing semantic information.
Related Papers
- An overview of text-to-speech synthesis techniques(2010)
- → An Overview of Speech Synthesis Technology(2018)7 cited
- → Ten Fun Word‐Building Activities(2008)
- → Faculty Opinions recommendation of Statistical laws governing fluctuations in word use from word birth to word death.(2012)
- → ҚАЗАҚ ТІЛІ ҮШІН ИНТЕГРАЛДЫҚ (END-TO-END) СӨЙЛЕУ СИНТЕЗІ(2023)