arXiv subjects

Lei Xie

Publications and source records attributed to Lei Xie.

At least 469 records · Page 26Linked to original sources

arXiv2020

Wake Word Detection with Alignment-Free Lattice-Free MMI

Always-on spoken language interfaces, e.g. personal digital assistants, rely on a wake word to start processing spoken input. We present novel methods to train a hybrid DNN/HMM wake word detection system from partially labeled training data, and to use it in on-line applications: (i) we remove the prerequisite of frame-level alignments in the LF-MMI training algorithm, permitting the use of un-transcribed training examples that are annotated only for the presence/absence of the wake word; (ii) we show that the classical keyword/filler model must be supplemented with an explicit non-speech (silence) model for good performance; (iii) we present an FST-based decoder to perform online detection. We evaluate our methods on two real data sets, showing 50%--90% reduction in false rejection rates at pre-specified false alarm rates over the best previously published figures, and re-validate them on a third (large) data set.

eess.AS↗

arXiv2020

The ASRU 2019 Mandarin-English Code-Switching Speech Recognition Challenge: Open Datasets, Tracks, Methods and Results

Code-switching (CS) is a common phenomenon and recognizing CS speech is challenging. But CS speech data is scarce and there' s no common testbed in relevant research. This paper describes the design and main outcomes of the ASRU 2019 Mandarin-English code-switching speech recognition challenge, which aims to improve the ASR performance in Mandarin-English code-switching situation. 500 hours Mandarin speech data and 240 hours Mandarin-English intra-sentencial CS data are released to the participants. Three tracks were set for advancing the AM and LM part in traditional DNN-HMM ASR system, as well as exploring the E2E models' performance. The paper then presents an overview of the results and system performance in the three tracks. It turns out that traditional ASR system benefits from pronunciation lexicon, CS text generating and data augmentation. In E2E track, however, the results highlight the importance of using language identification, building-up a rational set of modeling units and spec-augment. The other details in model training and method comparsion are discussed.

eess.AS↗

arXiv2020

Sequence to Multi-Sequence Learning via Conditional Chain Mapping for Mixture Signals

Neural sequence-to-sequence models are well established for applications which can be cast as mapping a single input sequence into a single output sequence. In this work, we focus on one-to-many sequence transduction problems, such as extracting multiple sequential sources from a mixture sequence. We extend the standard sequence-to-sequence model to a conditional multi-sequence model, which explicitly models the relevance between multiple output sequences with the probabilistic chain rule. Based on this extension, our model can conditionally infer output sequences one-by-one by making use of both input and previously-estimated contextual output sequences. This model additionally has a simple and efficient stop criterion for the end of the transduction, making it able to infer the variable number of output sequences. We take speech data as a primary test field to evaluate our methods since the observed speech data is often composed of multiple sources due to the nature of the superposition principle of sound waves. Experiments on several different tasks including speech separation and multi-speaker speech recognition show that our conditional multi-sequence models lead to consistent improvements over the conventional non-conditional models.

eess.AS↗

arXiv2020

Inaudible Adversarial Perturbations for Targeted Attack in Speaker Recognition

Speaker recognition is a popular topic in biometric authentication and many deep learning approaches have achieved extraordinary performances. However, it has been shown in both image and speech applications that deep neural networks are vulnerable to adversarial examples. In this study, we aim to exploit this weakness to perform targeted adversarial attacks against the x-vector based speaker recognition system. We propose to generate inaudible adversarial perturbations achieving targeted white-box attacks to speaker recognition system based on the psychoacoustic principle of frequency masking. Specifically, we constrict the perturbation under the masking threshold of original audio, instead of using a common l_p norm to measure the perturbations. Experiments on Aishell-1 corpus show that our approach yields up to 98.5% attack success rate to arbitrary gender speaker targets, while retaining indistinguishable attribute to listeners. Furthermore, we also achieve an effective speaker attack when applying the proposed approach to a completely irrelevant waveform, such as music.

cs.SD↗

arXiv2020

Streaming Chunk-Aware Multihead Attention for Online End-to-End Speech Recognition

Recently, streaming end-to-end automatic speech recognition (E2E-ASR) has gained more and more attention. Many efforts have been paid to turn the non-streaming attention-based E2E-ASR system into streaming architecture. In this work, we propose a novel online E2E-ASR system by using Streaming Chunk-Aware Multihead Attention(SCAMA) and a latency control memory equipped self-attention network (LC-SAN-M). LC-SAN-M uses chunk-level input to control the latency of encoder. As to SCAMA, a jointly trained predictor is used to control the output of encoder when feeding to decoder, which enables decoder to generate output in streaming manner. Experimental results on the open 170-hour AISHELL-1 and an industrial-level 20000-hour Mandarin speech recognition tasks show that our approach can significantly outperform the MoChA-based baseline system under comparable setup. On the AISHELL-1 task, our proposed method achieves a character error rate (CER) of 7.39%, to the best of our knowledge, which is the best published performance for online ASR.

cs.SD↗

arXiv2020

Improving Attention Mechanism in Graph Neural Networks via Cardinality Preservation

Graph Neural Networks (GNNs) are powerful to learn the representation of graph-structured data. Most of the GNNs use the message-passing scheme, where the embedding of a node is iteratively updated by aggregating the information of its neighbors. To achieve a better expressive capability of node influences, attention mechanism has grown to be popular to assign trainable weights to the nodes in aggregation. Though the attention-based GNNs have achieved remarkable results in various tasks, a clear understanding of their discriminative capacities is missing. In this work, we present a theoretical analysis of the representational properties of the GNN that adopts the attention mechanism as an aggregator. Our analysis determines all cases when those attention-based GNNs can always fail to distinguish certain distinct structures. Those cases appear due to the ignorance of cardinality information in attention-based aggregation. To improve the performance of attention-based GNNs, we propose cardinality preserved attention (CPA) models that can be applied to any kind of attention mechanisms. Our experiments on node and graph classification confirm our theoretical analysis and show the competitive performance of our CPA models.

cs.LG↗

arXiv2020

Adversarial Feature Learning and Unsupervised Clustering based Speech Synthesis for Found Data with Acoustic and Textual Noise

Attention-based sequence-to-sequence (seq2seq) speech synthesis has achieved extraordinary performance. But a studio-quality corpus with manual transcription is necessary to train such seq2seq systems. In this paper, we propose an approach to build high-quality and stable seq2seq based speech synthesis system using challenging found data, where training speech contains noisy interferences (acoustic noise) and texts are imperfect speech recognition transcripts (textual noise). To deal with text-side noise, we propose a VQVAE based heuristic method to compensate erroneous linguistic feature with phonetic information learned directly from speech. As for the speech-side noise, we propose to learn a noise-independent feature in the auto-regressive decoder through adversarial training and data augmentation, which does not need an extra speech enhancement model. Experiments show the effectiveness of the proposed approach in dealing with text-side and speech-side noise. Surpassing the denoising approach based on a state-of-the-art speech enhancement model, our system built on noisy found data can synthesize clean and high-quality speech with MOS close to the system built on the clean counterpart.

cs.SD↗

arXiv2020

RF-Rhythm: Secure and Usable Two-Factor RFID Authentication

Passive RFID technology is widely used in user authentication and access control. We propose RF-Rhythm, a secure and usable two-factor RFID authentication system with strong resilience to lost/stolen/cloned RFID cards. In RF-Rhythm, each legitimate user performs a sequence of taps on his/her RFID card according to a self-chosen secret melody. Such rhythmic taps can induce phase changes in the backscattered signals, which the RFID reader can detect to recover the user's tapping rhythm. In addition to verifying the RFID card's identification information as usual, the backend server compares the extracted tapping rhythm with what it acquires in the user enrollment phase. The user passes authentication checks if and only if both verifications succeed. We also propose a novel phase-hopping protocol in which the RFID reader emits Continuous Wave (CW) with random phases for extracting the user's secret tapping rhythm. Our protocol can prevent a capable adversary from extracting and then replaying a legitimate tapping rhythm from sniffed RFID signals. Comprehensive user experiments confirm the high security and usability of RF-Rhythm with false-positive and false-negative rates close to zero.

eess.SP↗

arXiv2019

a novel cross-lingual voice cloning approach with a few text-free samples

In this paper, we present a cross-lingual voice cloning approach. BN features obtained by SI-ASR model are used as a bridge across speakers and language boundaries. The relationships between text and BN features are modeled by the latent prosody model. The acoustic model learns the translation from BN features to acoustic features. The acoustic model is fine-tuned with a few samples of the target speaker to realize voice cloning. This system can generate speech of arbitrary utterance of target language in cross-lingual speakers' voice. We verify that with small amount of audio data, our proposed approach can well handle cross-lingual tasks. And in intra-lingual tasks, our proposed approach also performs better than baseline approach in naturalness and similarity.

eess.AS↗

arXiv2019

Espresso: A Fast End-to-end Neural Speech Recognition Toolkit

We present Espresso, an open-source, modular, extensible end-to-end neural automatic speech recognition (ASR) toolkit based on the deep learning library PyTorch and the popular neural machine translation toolkit fairseq. Espresso supports distributed training across GPUs and computing nodes, and features various decoding approaches commonly employed in ASR, including look-ahead word-based language model fusion, for which a fast, parallelized decoder is implemented. Espresso achieves state-of-the-art ASR performance on the WSJ, LibriSpeech, and Switchboard data sets among other end-to-end systems without data augmentation, and is 4--11x faster for decoding than similar systems (e.g. ESPnet).

cs.CL↗

arXiv2019

Time Domain Audio Visual Speech Separation

Audio-visual multi-modal modeling has been demonstrated to be effective in many speech related tasks, such as speech recognition and speech enhancement. This paper introduces a new time-domain audio-visual architecture for target speaker extraction from monaural mixtures. The architecture generalizes the previous TasNet (time-domain speech separation network) to enable multi-modal learning and at meanwhile it extends the classical audio-visual speech separation from frequency-domain to time-domain. The main components of proposed architecture include an audio encoder, a video encoder that extracts lip embedding from video streams, a multi-modal separation network and an audio decoder. Experiments on simulated mixtures based on recently released LRS2 dataset show that our method can bring 3dB+ and 4dB+ Si-SNR improvements on two- and three-speaker cases respectively, compared to audio-only TasNet and frequency-domain audio-visual networks

eess.AS↗

arXiv2019

Building a mixed-lingual neural TTS system with only monolingual data

When deploying a Chinese neural text-to-speech (TTS) synthesis system, one of the challenges is to synthesize Chinese utterances with English phrases or words embedded. This paper looks into the problem in the encoder-decoder framework when only monolingual data from a target speaker is available. Specifically, we view the problem from two aspects: speaker consistency within an utterance and naturalness. We start the investigation with an Average Voice Model which is built from multi-speaker monolingual data, i.e. Mandarin and English data. On the basis of that, we look into speaker embedding for speaker consistency within an utterance and phoneme embedding for naturalness and intelligibility and study the choice of data for model training. We report the findings and discuss the challenges to build a mixed-lingual TTS system with only monolingual data.

cs.CL↗

arXiv2019

Memristive Devices for Computation-In-Memory

CMOS technology and its continuous scaling have made electronics and computers accessible and affordable for almost everyone on the globe; in addition, they have enabled the solutions of a wide range of societal problems and applications. Today, however, both the technology and the computer architectures are facing severe challenges/walls making them incapable of providing the demanded computing power with tight constraints. This motivates the need for the exploration of novel architectures based on new device technologies; not only to sustain the financial benefit of technology scaling, but also to develop solutions for extremely demanding emerging applications. This paper presents two computation-in-memory based accelerators making use of emerging memristive devices; they are Memristive Vector Processor and RRAM Automata Processor. The preliminary results of these two accelerators show significant improvement in terms of latency, energy and area as compared to today's architectures and design.

cs.ET↗

arXiv2019

Exploring RNN-Transducer for Chinese Speech Recognition

End-to-end approaches have drawn much attention recently for significantly simplifying the construction of an automatic speech recognition (ASR) system. RNN transducer (RNN-T) is one of the popular end-to-end methods. Previous studies have shown that RNN-T is difficult to train and a very complex training process is needed for a reasonable performance. In this paper, we explore RNN-T for a Chinese large vocabulary continuous speech recognition (LVCSR) task and aim to simplify the training process while maintaining performance. First, a new strategy of learning rate decay is proposed to accelerate the model convergence. Second, we find that adding convolutional layers at the beginning of the network and using ordered data can discard the pre-training process of the encoder without loss of performance. Besides, we design experiments to find a balance among the usage of GPU memory, training circle and model performance. Finally, we achieve 16.9% character error rate (CER) on our test set which is 2% absolute improvement from a strong BLSTM CE system with language model trained on the same text corpus.

cs.CL↗

arXiv2019

Exploiting Syntactic Features in a Parsed Tree to Improve End-to-End TTS

The end-to-end TTS, which can predict speech directly from a given sequence of graphemes or phonemes, has shown improved performance over the conventional TTS. However, its predicting capability is still limited by the acoustic/phonetic coverage of the training data, usually constrained by the training set size. To further improve the TTS quality in pronunciation, prosody and perceived naturalness, we propose to exploit the information embedded in a syntactically parsed tree where the inter-phrase/word information of a sentence is organized in a multilevel tree structure. Specifically, two key features: phrase structure and relations between adjacent words are investigated. Experimental results in subjective listening, measured on three test sets, show that the proposed approach is effective to improve the pronunciation clarity, prosody and naturalness of the synthesized speech of the baseline system.

cs.CL↗

arXiv2019

A New GAN-based End-to-End TTS Training Algorithm

End-to-end, autoregressive model-based TTS has shown significant performance improvements over the conventional one. However, the autoregressive module training is affected by the exposure bias, or the mismatch between the different distributions of real and predicted data. While real data is available in training, but in testing, only predicted data is available to feed the autoregressive module. By introducing both real and generated data sequences in training, we can alleviate the effects of the exposure bias. We propose to use Generative Adversarial Network (GAN) along with the key idea of Professor Forcing in training. A discriminator in GAN is jointly trained to equalize the difference between real and predicted data. In AB subjective listening test, the results show that the new approach is preferred over the standard transfer learning with a CMOS improvement of 0.1. Sentence level intelligibility tests show significant improvement in a pathological test set. The GAN-trained new model is also more stable than the baseline to produce better alignments for the Tacotron output.

cs.CL↗

arXiv2019

Improved Speaker-Dependent Separation for CHiME-5 Challenge

This paper summarizes several follow-up contributions for improving our submitted NWPU speaker-dependent system for CHiME-5 challenge, which aims to solve the problem of multi-channel, highly-overlapped conversational speech recognition in a dinner party scenario with reverberations and non-stationary noises. We adopt a speaker-aware training method by using i-vector as the target speaker information for multi-talker speech separation. With only one unified separation model for all speakers, we achieve a 10\% absolute improvement in terms of word error rate (WER) over the previous baseline of 80.28\% on the development set by leveraging our newly proposed data processing techniques and beamforming approach. With our improved back-end acoustic model, we further reduce WER to 60.15\% which surpasses the result of our submitted CHiME-5 challenge system without applying any fusion techniques.

eess.AS↗

arXiv2018

Structural insights into characterizing binding sites in EGFR kinase mutants

Over the last two decades epidermal growth factor receptor (EGFR) kinase has become an important target to treat non-small cell lung cancer (NSCLC). Currently, three generations of EGFR kinase-targeted small molecule drugs have been FDA approved. They nominally produce a response at the start of treatment and lead to a substantial survival benefit for patients. However, long-term treatment results in acquired drug resistance and further vulnerability to NSCLC. Therefore, novel EGFR kinase inhibitors that specially overcome acquired mutations are urgently needed. To this end, we carried out a comprehensive study of different EGFR kinase mutants using a structural systems pharmacology strategy. Our analysis shows that both wild-type and mutated structures exhibit multiple conformational states that have not been observed in solved crystal structures. We show that this conformational flexibility accommodates diverse types of ligands with multiple types of binding modes. These results provide insights for designing a new-generation of EGFR kinase inhibitor that combats acquired drug-resistant mutations through a multi-conformation-based drug design strategy.

q-bio.MN↗