Zobrazeno 1 - 10
of 561
pro vyhledávání: '"Wu Felix"'
This work introduces Sample-Efficient Speech Diffusion (SESD), an algorithm for effective speech synthesis in modest data regimes through latent diffusion. It is based on a novel diffusion architecture, that we call U-Audio Transformer (U-AT), that e
Externí odkaz:
http://arxiv.org/abs/2409.03717
Autor:
Cooper, A. Feder, Lee, Katherine, Grimmelmann, James, Ippolito, Daphne, Callison-Burch, Christopher, Choquette-Choo, Christopher A., Mireshghallah, Niloofar, Brundage, Miles, Mimno, David, Choksi, Madiha Zahrah, Balkin, Jack M., Carlini, Nicholas, De Sa, Christopher, Frankle, Jonathan, Ganguli, Deep, Gipson, Bryant, Guadamuz, Andres, Harris, Swee Leng, Jacobs, Abigail Z., Joh, Elizabeth, Kamath, Gautam, Lemley, Mark, Matthews, Cass, McLeavey, Christine, McSherry, Corynne, Nasr, Milad, Ohm, Paul, Roberts, Adam, Rubin, Tom, Samuelson, Pamela, Schubert, Ludwig, Vaccaro, Kristen, Villa, Luis, Wu, Felix, Zeide, Elana
This report presents the takeaways of the inaugural Workshop on Generative AI and Law (GenLaw), held in July 2023. A cross-disciplinary group of practitioners and scholars from computer science and law convened to discuss the technical, doctrinal, an
Externí odkaz:
http://arxiv.org/abs/2311.06477
A common training technique for language models is teacher forcing (TF). TF attempts to match human language exactly, even though identical meanings can be expressed in different ways. This motivates use of sequence-level objectives for dialogue resp
Externí odkaz:
http://arxiv.org/abs/2307.12425
Autor:
Peng, Yifan, Kim, Kwangyoun, Wu, Felix, Yan, Brian, Arora, Siddhant, Chen, William, Tang, Jiyang, Shon, Suwon, Sridhar, Prashant, Watanabe, Shinji
Conformer, a convolution-augmented Transformer variant, has become the de facto encoder architecture for speech processing due to its superior performance in various tasks, including automatic speech recognition (ASR), speech translation (ST) and spo
Externí odkaz:
http://arxiv.org/abs/2305.11073
Self-supervised speech representation learning (SSL) has shown to be effective in various downstream tasks, but SSL models are usually large and slow. Model compression techniques such as pruning aim to reduce the model size and computation without d
Externí odkaz:
http://arxiv.org/abs/2302.14132
Autor:
Shon, Suwon, Arora, Siddhant, Lin, Chyi-Jiunn, Pasad, Ankita, Wu, Felix, Sharma, Roshan, Wu, Wei-Lun, Lee, Hung-Yi, Livescu, Karen, Watanabe, Shinji
Spoken language understanding (SLU) tasks have been studied for many decades in the speech research community, but have not received as much attention as lower-level tasks like speech and speaker recognition. In particular, there are not nearly as ma
Externí odkaz:
http://arxiv.org/abs/2212.10525
Self-supervised pre-trained transformers have improved the state of the art on a variety of speech tasks. Due to the quadratic time and space complexity of self-attention, they usually operate at the level of relatively short (e.g., utterance) segmen
Externí odkaz:
http://arxiv.org/abs/2212.08542
Autor:
Kim, Kwangyoun, Wu, Felix, Peng, Yifan, Pan, Jing, Sridhar, Prashant, Han, Kyu J., Watanabe, Shinji
Conformer, combining convolution and self-attention sequentially to capture both local and global information, has shown remarkable performance and is currently regarded as the state-of-the-art for automatic speech recognition (ASR). Several other st
Externí odkaz:
http://arxiv.org/abs/2210.00077
Autor:
Wu, Felix, Kim, Kwangyoun, Watanabe, Shinji, Han, Kyu, McDonald, Ryan, Weinberger, Kilian Q., Artzi, Yoav
We introduce Wav2Seq, the first self-supervised approach to pre-train both parts of encoder-decoder models for speech data. We induce a pseudo language as a compact discrete representation, and formulate a self-supervised pseudo speech recognition ta
Externí odkaz:
http://arxiv.org/abs/2205.01086
Spoken language understanding (SLU) tasks involve mapping from speech audio signals to semantic labels. Given the complexity of such tasks, good performance might be expected to require large labeled datasets, which are difficult to collect for each
Externí odkaz:
http://arxiv.org/abs/2112.07648