Zobrazeno 1 - 10
of 21
pro vyhledávání: '"Wu, Chunyang"'
We introduce Speech ReaLLM, a new ASR architecture that marries "decoder-only" ASR with the RNN-T to make multimodal LLM architectures capable of real-time streaming. This is the first "decoder-only" ASR architecture designed to handle continuous aud
Externí odkaz:
http://arxiv.org/abs/2406.09569
Autor:
Guo, Jinxi, Moritz, Niko, Ma, Yingyi, Seide, Frank, Wu, Chunyang, Mahadeokar, Jay, Kalinli, Ozlem, Fuegen, Christian, Seltzer, Mike
The internal language model (ILM) of the neural transducer has been widely studied. In most prior work, it is mainly used for estimating the ILM score and is subsequently subtracted during inference to facilitate improved integration with external la
Externí odkaz:
http://arxiv.org/abs/2404.01716
Autor:
Xie, Jiamin, Li, Ke, Guo, Jinxi, Tjandra, Andros, Shangguan, Yuan, Sari, Leda, Wu, Chunyang, Jia, Junteng, Mahadeokar, Jay, Kalinli, Ozlem
Neural network pruning offers an effective method for compressing a multilingual automatic speech recognition (ASR) model with minimal performance loss. However, it entails several rounds of pruning and re-training needed to be run for each language.
Externí odkaz:
http://arxiv.org/abs/2309.13018
Autor:
Lakomkin, Egor, Wu, Chunyang, Fathullah, Yassir, Kalinli, Ozlem, Seltzer, Michael L., Fuegen, Christian
In recent years, Large Language Models (LLMs) have garnered significant attention from the research community due to their exceptional performance and generalization capabilities. In this paper, we introduce a novel method for contextualizing speech
Externí odkaz:
http://arxiv.org/abs/2309.10917
Autor:
Shangguan, Yuan, Yang, Haichuan, Li, Danni, Wu, Chunyang, Fathullah, Yassir, Wang, Dilin, Dalmia, Ayushi, Krishnamoorthi, Raghuraman, Kalinli, Ozlem, Jia, Junteng, Mahadeokar, Jay, Lei, Xin, Seltzer, Mike, Chandra, Vikas
Automatic Speech Recognition (ASR) models need to be optimized for specific hardware before they can be deployed on devices. This can be done by tuning the model's hyperparameters or exploring variations in its architecture. Re-training and re-valida
Externí odkaz:
http://arxiv.org/abs/2309.01947
Autor:
Fathullah, Yassir, Wu, Chunyang, Lakomkin, Egor, Jia, Junteng, Shangguan, Yuan, Li, Ke, Guo, Jinxi, Xiong, Wenhan, Mahadeokar, Jay, Kalinli, Ozlem, Fuegen, Christian, Seltzer, Mike
Large language models have proven themselves highly flexible, able to solve a wide range of generative tasks, such as abstractive summarization and open-ended question answering. In this paper we extend the capabilities of LLMs by directly attaching
Externí odkaz:
http://arxiv.org/abs/2307.11795
Autor:
Liu, Shuo, Sarı, Leda, Wu, Chunyang, Keren, Gil, Shangguan, Yuan, Mahadeokar, Jay, Kalinli, Ozlem
This paper presents a method for selecting appropriate synthetic speech samples from a given large text-to-speech (TTS) dataset as supplementary training data for an automatic speech recognition (ASR) model. We trained a neural network, which can be
Externí odkaz:
http://arxiv.org/abs/2306.00998
Autor:
Fathullah, Yassir, Wu, Chunyang, Shangguan, Yuan, Jia, Junteng, Xiong, Wenhan, Mahadeokar, Jay, Liu, Chunxi, Shi, Yangyang, Kalinli, Ozlem, Seltzer, Mike, Gales, Mark J. F.
State space models (SSMs) have recently shown promising results on small-scale sequence and language modelling tasks, rivalling and outperforming many attention-based approaches. In this paper, we propose a multi-head state space (MH-SSM) architectur
Externí odkaz:
http://arxiv.org/abs/2305.12498
Autor:
Raj, Desh, Jia, Junteng, Mahadeokar, Jay, Wu, Chunyang, Moritz, Niko, Zhang, Xiaohui, Kalinli, Ozlem
Neural transducers have achieved human level performance on standard speech recognition benchmarks. However, their performance significantly degrades in the presence of cross-talk, especially when the primary speaker has a low signal-to-noise ratio.
Externí odkaz:
http://arxiv.org/abs/2210.11588
Autor:
Shi, Yangyang, Wu, Chunyang, Wang, Dilin, Xiao, Alex, Mahadeokar, Jay, Zhang, Xiaohui, Liu, Chunxi, Li, Ke, Shangguan, Yuan, Nagaraja, Varun, Kalinli, Ozlem, Seltzer, Mike
This paper improves the streaming transformer transducer for speech recognition by using non-causal convolution. Many works apply the causal convolution to improve streaming transformer ignoring the lookahead context. We propose to use non-causal con
Externí odkaz:
http://arxiv.org/abs/2110.05241