SpecAugment: A New Data Augmentation Method for Automatic Speech Recognition – Google Research Blog

Posted by Daniel S. Park, AI Resident and William Chan, Research Scientist

SpecAugment: A New Data Augmentation Method for Automatic Speech Recognition Skip to main content SpecAugment: A New Data Augmentation Method for Automatic Speech Recognition April 22, 2019 Posted by Daniel S. Park, AI Resident and William Chan, Research Scientist Quick links Share Copy link × Automatic Speech Recognition (ASR), the process of taking an audio input and transcribing it to text, has benefited greatly from the ongoing development of deep neural networks . As a result, ASR has become ubiquitous in many modern devices and products, such as Google Assistant, Google Home and YouTube.

saved by

• Ryan Conti

related reading

• Audio Deep Learning Made Simple: Automatic Speech Recognition (ASR), How it Works | Towards Data Sciencetowardsdatascience.com
• An Intuitive Explanation of Sparse Autoencoders for LLM Interpretability | Adam Karvonenadamkarvonen.github.io
• Silent speech with ultrasound — Alephalephneuro.com
• An Open Base Model for Conversational Speech | blogsi.inc
• [2105.11084] Unsupervised Speech Recognitionarxiv.org
• [2005.11401] Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasksarxiv.org
• Sparse Autoencoders Find Highly Interpretable Features in Language Modelsarxiv.org
• The Annotated Transformernlp.seas.harvard.edu
• Sequence Modeling with CTCdistill.pub
• A Survey on Sparse Autoencoders: Interpreting the Internal Mechanisms of Large Language Modelsarxiv.org
• 2206.14483arxiv.org
• Augmented Audio Research | Augmented Audio Researchaugmented.audio