Summary of Joint Unsupervised and Supervised Training For Automatic Speech Recognition Via Bilevel Optimization, by a F M Saif et al.
Joint Unsupervised and Supervised Training for Automatic Speech Recognition via Bilevel Optimization
by A F M Saif, Xiaodong Cui, Han Shen, Songtao Lu, Brian Kingsbury, Tianyi Chen
First submitted to arxiv on: 13 Jan 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Machine Learning (cs.LG); Machine Learning (stat.ML)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This novel bilevel optimization-based training approach, dubbed {bi-level joint unsupervised and supervised training (BL-JUST)}, tackles automatic speech recognition (ASR) tasks with remarkable success. By combining an unsupervised loss and a supervised loss through penalty-based bilevel optimization, BL-JUST efficiently solves the challenging ASR problem while providing rigorous convergence guarantees. Experiments on LibriSpeech and TED-LIUM v2 datasets demonstrate that BL-JUST outperforms the traditional pre-training followed by fine-tuning strategy. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary BL-JUST is a new way to train acoustic models for automatic speech recognition. It’s like solving a puzzle with two pieces: one from unsupervised learning and another from supervised learning. The results show that this approach works better than usual methods on big datasets like LibriSpeech and TED-LIUM. |
Keywords
* Artificial intelligence * Fine tuning * Optimization * Supervised * Unsupervised
