Zobrazeno 1 - 10 of 55 pro vyhledávání: '"Cyril Allauzen"'
1
Large-scale Language Model Rescoring on Long-form Data
Autor: Tongzhou Chen, Cyril Allauzen, Yinghui Huang, Daniel Park, David Rybach, W. Ronny Huang, Rodrigo Cabrera, Kartik Audhkhasi, Bhuvana Ramabhadran, Pedro J. Moreno, Michael Riley
In this work, we study the impact of Large-scale Language Models (LLM) on Automated Speech Recognition (ASR) of YouTube videos, which we use as a source for long-form ASR. We demonstrate up to 8\% relative reduction in Word Error Eate (WER) on US Eng
Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::0d6983808ef866164081e3c5a2b3ffe8
http://arxiv.org/abs/2306.08133
Zobrazit plný text záznamu
3
E2E Segmenter: Joint Segmenting and Decoding for Long-Form ASR
Autor: W. Ronny Huang, Shuo-Yiin Chang, David Rybach, Tara Sainath, Rohit Prabhavalkar, Cal Peyser, Zhiyun Lu, Cyril Allauzen
Improving the performance of end-to-end ASR models on long utterances ranging from minutes to hours in length is an ongoing challenge in speech recognition. A common solution is to segment the audio in advance using a separate voice activity detector
Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::e9c15fa03f1db0e0b6fc4f4d4951f31a
Zobrazit plný text záznamu
5
Hybrid Autoregressive Transducer (hat)
Autor: David Rybach, Ehsan Variani, Michael Riley, Cyril Allauzen
Publikováno v: ICASSP
This paper proposes and evaluates the hybrid autoregressive transducer (HAT) model, a time-synchronous encoderdecoder model that preserves the modularity of conventional automatic speech recognition systems. The HAT model provides a way to measure th
Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::f20420d25b46848583b2c48806c80a4d
Zobrazit plný text záznamu
7
Federated Learning of N-gram Language Models
Autor: Francoise Beaufays, Adeline Wong, Michael Riley, Mingqing Chen, Ananda Theertha Suresh, Cyril Allauzen, Rajiv Mathews
Publikováno v: CoNLL
We propose algorithms to train production-quality n-gram language models using federated learning. Federated learning is a distributed computation platform that can be used to train global models for portable devices such as smart phones. Federated l
Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::fb9bc59794980ecf0370788af5ddfeac
Zobrazit plný text záznamu
8
On the Compression of Lexicon Transducers
Autor: Marco Cognetta, Cyril Allauzen, Michael A. Riley
Publikováno v: FSMNLP
In finite-state language processing pipelines, a lexicon is often a key component. It needs to be comprehensive to ensure accuracy, reducing out-of-vocabulary misses. However, in memory-constrained environments (e.g., mobile phones), the size of the
Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_________::8dfe334bb4802a49cb0fdba94c9c6c07
https://doi.org/10.18653/v1/w19-3105
Zobrazit plný text záznamu
9
Algorithms for Weighted Finite Automata with Failure Transitions
Autor: Cyril Allauzen, Michael Riley
Publikováno v: Implementation and Application of Automata ISBN: 9783319948119
CIAA
In this paper we extend several weighted finite automata (WFA) algorithms to automata with failure transitions (\(\varphi \)-WFAs). Failure transitions, which are taken only when no immediate match is possible at a given state, are used to compactly
Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_________::5f877bfa724bfa68ee9edb67ae49eb77
https://doi.org/10.1007/978-3-319-94812-6_5
Zobrazit plný text záznamu

Vyhledávací nástroje:

Upřesnit hledání

Omezení vyhledávání
Plný text Recenzováno Digitální knihovna AV ČR
Zdroje