Google Inc.

Baidu USA LLC

Jesse Engel

Adam Roberts

Bryan Catanzaro

Erich Elsen

Shubhabrata Sengupta

Awni Hannun

Dario Amodei

Matthew Nirvan Sharifi

Xu Han

Marco Tagliasacchi

Chenjie Gu

Ryan Prenger

Jun Zhan

Neil Zeghidour

Sanjeev Satheesh

Zhenyao Zhu

Christopher Fougner

Chong Wang

Zalán Borsos

Mike Chrzanowski

Christian Frank

Dani Yogatama

Jingdong Chen

Lamtharn Hantrakul

Mauro Verzetti

Antoine Caillon

Timo Immanuel Denk

Andrea Agostinelli

Karen Simonyan

Douglas Eck

Gregory Diamos

Mohammad Norouzi

Sander Etienne Lea Dieleman

Cinjon Resnick

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating a prediction of an audio signal. One of the methods includes receiving a request to generate an audio signal conditioned on an input; processing the input using an embedding neural network to map the input to one or more embedding tokens; generating a semantic representation of the audio signal; generating, using one or more generative neural networks and conditioned on at least the semantic representation and the embedding tokens, an acoustic representation of the audio signal; and processing at least the acoustic representation using a decoder neural network to generate the prediction of the audio signal.

Generating audio using auto-regressive generative neural networks

Systems and methods of the present disclosure are directed toward digital signal processing using machine-learned differentiable digital signal processors. For example, embodiments of the present disclosure may include differentiable digital signal processors within the training loop of a machine-learned model (e.g., for gradient-based training). Advantageously, systems and methods of the present disclosure provide high quality signal processing using smaller models than prior systems, thereby reducing energy costs (e.g., storage and/or processing costs) associated with performing digital signal processing.

Machine-learned differentiable digital signal processing

Systems and methods for a multi-core optimized Recurrent Neural Network (RNN) architecture are disclosed. The various architectures affect communication and synchronization operations according to the Multi-Bulk-Synchronous-Parallel (MBSP) model for a given processor. The resulting family of network architectures, referred to as MBSP-RNNs, perform similarly to a conventional RNNs having the same number of parameters, but are substantially more efficient when mapped onto a modern general purpose processor. Due to the large gain in computational efficiency, for a fixed computational budget, MBSP-RNNs outperform RNNs at applications such as end-to-end speech recognition.

Systems and methods for a multi-core optimized recurrent neural network

Embodiments of end-to-end deep learning systems and methods are disclosed to recognize speech of vastly different languages, such as English or Mandarin Chinese. In embodiments, the entire pipelines of hand-engineered components are replaced with neural networks, and the end-to-end learning allows handling a diverse variety of speech including noisy environments, accents, and different languages. Using a trained embodiment and an embodiment of a batch dispatch technique with GPUs in a data center, an end-to-end deep learning system can be inexpensively deployed in an online setting, delivering low latency when serving users at scale.

End-to-end speech recognition

Deployed end-to-end speech recognition

The present disclosure provides systems and methods that include or otherwise leverage a machine-learned neural synthesizer model. Unlike a traditional synthesizer which generates audio from hand-designed components like oscillators and wavetables, the neural synthesizer model can use deep neural networks to generate sounds at the level of individual samples. Learning directly from data, the neural synthesizer model can provide intuitive control over timbre and dynamics and enable exploration of new sounds that would be difficult or impossible to produce with a hand-tuned synthesizer. As one example, the neural synthesizer model can be a neural synthesis autoencoder that includes an encoder model that learns embeddings descriptive of musical characteristics and an autoregressive decoder model that is conditioned on the embedding to autoregressively generate musical waveforms that have the musical characteristics one audio sample at a time.

Generating music with deep neural networks

Growing community of inventors

Oakland, CA, United States of America

Jesse Engel