Summary of Amps: Asr with Multimodal Paraphrase Supervision, by Amruta Parulekar et al.
AMPS: ASR with Multimodal Paraphrase Supervision
by Amruta Parulekar, Abhishek Gupta, Sameep Chattopadhyay, Preethi Jyothi
First submitted to arxiv on: 27 Nov 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Audio and Speech Processing (eess.AS)
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 The proposed technique AMPS augments a multilingual multimodal ASR system with paraphrase-based supervision for improved conversational ASR in multiple languages. The approach uses paraphrases of reference transcriptions as additional supervision during training, selectively invoking this objective for utterances with poor ASR performance. By combining AMPS with the state-of-the-art SeamlessM4T model, significant reductions in word error rates (WERs) of up to 5% are achieved. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper makes it easier for machines to understand spoken language from different countries and cultures. It develops a new way to improve automatic speech recognition systems that can handle conversations in multiple languages. This is helpful because many people speak multiple languages, but current technology struggles to keep up with the nuances of these conversations. The new method uses special examples of correct transcriptions as clues to help the system learn. |
