477 citations · 1.7k across the 21 of their papers we have counts for
22 papers
Kaizen: Continuously improving teacher using Exponential Moving Average for semi-supervised speech recognition
Vimal Manohar, Tatiana Likhomanenko, Qiantong Xu +5
In this paper, we introduce the Kaizen framework that uses a continuously improving teacher to generate pseudo-labels for semi-supervised speech recognition (ASR). The proposed app…
Benchmarking LF-MMI, CTC and RNN-T Criteria for Streaming ASR
Xiaohui Zhang, Frank Zhang, Chunxi Liu +8
In this work, to measure the accuracy and efficiency for a latency-controlled streaming automatic speech recognition (ASR) application, we perform comprehensive evaluations on thre…
Improving RNN Transducer Based ASR with Auxiliary Tasks
Chunxi Liu, Frank Zhang, Duc Le +3
End-to-end automatic speech recognition (ASR) models with a single neural network have recently demonstrated state-of-the-art results compared to conventional hybrid speech recogni…
Contextualizing ASR Lattice Rescoring with Hybrid Pointer Network Language Model
Da-Rong Liu, Chunxi Liu, Frank Zhang +3
Videos uploaded on social media are often accompanied with textual descriptions. In building automatic speech recognition (ASR) systems for videos, we can exploit the contextual in…
Faster, Simpler and More Accurate Hybrid ASR Systems Using Wordpieces
Frank Zhang, Yongqiang Wang, Xiaohui Zhang +3
In this work, we first show that on the widely used LibriSpeech benchmark, our transformer-based context-dependent connectionist temporal classification (CTC) system produces state…
Large scale weakly and semi-supervised learning for low-resource video ASR
Kritika Singh, Vimal Manohar, Alex Xiao +7
Many semi- and weakly-supervised approaches have been investigated for overcoming the labeling cost of building high quality speech recognition systems. On the challenging task of…
On Compositions of Transformations in Contrastive Self-Supervised Learning
Mandela Patrick, Yuki M. Asano, Polina Kuznetsova +4
In the image domain, excellent representations can be learned by inducing invariance to content-preserving transformations via noise contrastive learning. In this paper, we general…
From Senones to Chenones: Tied Context-Dependent Graphemes for Hybrid Speech Recognition
Duc Le, Xiaohui Zhang, Weiyi Zheng +3
There is an implicit assumption that traditional hybrid approaches for automatic speech recognition (ASR) cannot directly model graphemes and need to rely on phonetic lexicons to g…
Training ASR models by Generation of Contextual Information
Kritika Singh, Dmytro Okhonko, Jun Liu +8
Supervised ASR models have reached unprecedented levels of accuracy, thanks in part to ever-increasing amounts of labelled training data. However, in many applications and locales,…
Deja-vu: Double Feature Presentation and Iterated Loss in Deep Transformer Networks
Andros Tjandra, Chunxi Liu, Frank Zhang +5
Deep acoustic models typically receive features in the first layer of the network, and process increasingly abstract representations in the subsequent layers. Here, we propose to f…
Transformer-based Acoustic Modeling for Hybrid Speech Recognition
Yongqiang Wang, Abdelrahman Mohamed, Duc Le +10
We propose and evaluate transformer-based acoustic models (AMs) for hybrid speech recognition. Several modeling choices are discussed in this work, including various positional emb…
Multilingual Graphemic Hybrid ASR with Massive Data Augmentation
Chunxi Liu, Qiaochu Zhang, Xiaohui Zhang +3
Towards developing high-performing ASR for low-resource languages, approaches to address the lack of resources are to make use of data from multiple languages, and to augment the t…
Hybrid Code Networks: practical and efficient end-to-end dialog control with supervised and reinforcement learning
Jason D. Williams, Kavosh Asadi, Geoffrey Zweig
End-to-end learning of recurrent neural networks (RNNs) is an attractive solution for dialog systems; however, current techniques are data-intensive and require thousands of dialog…
Achieving Human Parity in Conversational Speech Recognition
W. Xiong, J. Droppo, X. Huang +5
Conversational speech recognition has served as a flagship speech recognition task since the release of the Switchboard corpus in the 1990s. In this paper, we measure the human err…
Advances in All-Neural Speech Recognition
G. Zweig, C. Yu, J. Droppo +1
This paper advances the design of CTC-based all-neural (or end-to-end) speech recognizers. We propose a novel symbol inventory, and a novel iterated-CTC method in which a second sy…
The Microsoft 2016 Conversational Speech Recognition System
W. Xiong, J. Droppo, X. Huang +5
We describe Microsoft's conversational speech recognition system, in which we combine recent developments in neural-network-based acoustic and language modeling to advance the stat…
An Attentional Neural Conversation Model with Improved Specificity
Kaisheng Yao, Baolin Peng, Geoffrey Zweig +1
In this paper we propose a neural conversation model for conducting dialogues. We demonstrate the use of this model to generate help desk responses, where users are asking question…
End-to-end LSTM-based dialog control optimized with supervised and reinforcement learning
Jason D. Williams, Geoffrey Zweig
This paper presents a model for end-to-end learning of task-oriented dialog systems. The main component of the model is a recurrent neural network (an LSTM), which maps from raw di…
Attention with Intention for a Neural Network Conversation Model
Kaisheng Yao, Geoffrey Zweig, Baolin Peng
In a conversation or a dialogue process, attention and intention play intrinsic roles. This paper proposes a neural network based approach that models the attention and intention p…
Sequence-to-Sequence Neural Net Models for Grapheme-to-Phoneme Conversion
Kaisheng Yao, Geoffrey Zweig
Sequence-to-sequence translation methods based on generation with a side-conditioned language model have recently shown promising results in several tasks. In machine translation,…
Language Models for Image Captioning: The Quirks and What Works
Jacob Devlin, Hao Cheng, Hao Fang +5
Two recent approaches have achieved state-of-the-art results in image captioning. The first uses a pipelined process where a set of candidate words is generated by a convolutional…
From Captions to Visual Concepts and Back
Hao Fang, Saurabh Gupta, Forrest Iandola +9
This paper presents a novel approach for automatically generating image descriptions: visual detectors, language models, and multimodal similarity models learnt directly from a dat…