Improved Contextual Recognition In Automatic Speech Recognition Systems By Semantic Lattice Rescoring
This work addresses the challenge of contextual recognition in ASR systems, which is crucial for improving transcription accuracy in conversational agents, but it appears incremental as it builds on existing HMM-GMM and DNN methods.
The paper tackled the problem of accurately discerning context-dependent words and phrases in automatic speech recognition by proposing a semantic lattice rescoring approach using transformer-based models, resulting in a palpable reduction in Word Error Rate on the LibriSpeech dataset.
Automatic Speech Recognition (ASR) has witnessed a profound research interest. Recent breakthroughs have given ASR systems different prospects such as faithfully transcribing spoken language, which is a pivotal advancement in building conversational agents. However, there is still an imminent challenge of accurately discerning context-dependent words and phrases. In this work, we propose a novel approach for enhancing contextual recognition within ASR systems via semantic lattice processing leveraging the power of deep learning models in accurately delivering spot-on transcriptions across a wide variety of vocabularies and speaking styles. Our solution consists of using Hidden Markov Models and Gaussian Mixture Models (HMM-GMM) along with Deep Neural Networks (DNN) models integrating both language and acoustic modeling for better accuracy. We infused our network with the use of a transformer-based model to properly rescore the word lattice achieving remarkable capabilities with a palpable reduction in Word Error Rate (WER). We demonstrate the effectiveness of our proposed framework on the LibriSpeech dataset with empirical analyses.