arXiv subjects

Xiaofei Wang

Publications and source records attributed to Xiaofei Wang.

At least 163 records · Page 9Linked to original sources

arXiv2019

Multi-Stream End-to-End Speech Recognition

Attention-based methods and Connectionist Temporal Classification (CTC) network have been promising research directions for end-to-end (E2E) Automatic Speech Recognition (ASR). The joint CTC/Attention model has achieved great success by utilizing both architectures during multi-task training and joint decoding. In this work, we present a multi-stream framework based on joint CTC/Attention E2E ASR with parallel streams represented by separate encoders aiming to capture diverse information. On top of the regular attention networks, the Hierarchical Attention Network (HAN) is introduced to steer the decoder toward the most informative encoders. A separate CTC network is assigned to each stream to force monotonic alignments. Two representative framework have been proposed and discussed, which are Multi-Encoder Multi-Resolution (MEM-Res) framework and Multi-Encoder Multi-Array (MEM-Array) framework, respectively. In MEM-Res framework, two heterogeneous encoders with different architectures, temporal resolutions and separate CTC networks work in parallel to extract complimentary information from same acoustics. Experiments are conducted on Wall Street Journal (WSJ) and CHiME-4, resulting in relative Word Error Rate (WER) reduction of 18.0-32.1% and the best WER of 3.6% in the WSJ eval92 test set. The MEM-Array framework aims at improving the far-field ASR robustness using multiple microphone arrays which are activated by separate encoders. Compared with the best single-array results, the proposed framework has achieved relative WER reduction of 3.7% and 9.7% in AMI and DIRHA multi-array corpora, respectively, which also outperforms conventional fusion strategies.

eess.AS↗

arXiv2019

DeshadowGAN: A Deep Learning Approach to Remove Shadows from Optical Coherence Tomography Images

Purpose: To remove retinal shadows from optical coherence tomography (OCT) images of the optic nerve head(ONH). Methods:2328 OCT images acquired through the center of the ONH using a Spectralis OCT machine for both eyes of 13 subjects were used to train a generative adversarial network (GAN) using a custom loss function. Image quality was assessed qualitatively (for artifacts) and quantitatively using the intralayer contrast: a measure of shadow visibility ranging from 0 (shadow-free) to 1 (strong shadow) and compared to compensated images. This was computed in the Retinal Nerve Fiber Layer (RNFL), the Inner Plexiform Layer (IPL), the Photoreceptor layer (PR) and the Retinal Pigment Epithelium (RPE) layers. Results: Output images had improved intralayer contrast in all ONH tissue layers. On average the intralayer contrast decreased by 33.7$\pm$6.81%, 28.8$\pm$10.4%, 35.9$\pm$13.0%, and43.0$\pm$19.5%for the RNFL, IPL, PR, and RPE layers respectively, indicating successful shadow removal across all depths. This compared to 70.3$\pm$22.7%, 33.9$\pm$11.5%, 47.0$\pm$11.2%, 26.7$\pm$19.0%for compensation. Output images were also free from artifacts commonly observed with compensation. Conclusions: DeshadowGAN significantly corrected blood vessel shadows in OCT images of the ONH. Our algorithm may be considered as a pre-processing step to improve the performance of a wide range of algorithms including those currently being used for OCT image segmentation, denoising, and classification. Translational Relevance: DeshadowGAN could be integrated to existing OCT devices to improve the diagnosis and prognosis of ocular pathologies.

eess.IV↗

arXiv2019

A Comparative Study on Transformer vs RNN in Speech Applications

Sequence-to-sequence models have been widely used in end-to-end speech processing, for example, automatic speech recognition (ASR), speech translation (ST), and text-to-speech (TTS). This paper focuses on an emergent sequence-to-sequence model called Transformer, which achieves state-of-the-art performance in neural machine translation and other natural language processing applications. We undertook intensive studies in which we experimentally compared and analyzed Transformer and conventional recurrent neural networks (RNN) in a total of 15 ASR, one multilingual ASR, one ST, and two TTS benchmarks. Our experiments revealed various training tips and significant performance benefits obtained with Transformer for each task including the surprising superiority of Transformer in 13/15 ASR benchmarks in comparison with RNN. We are preparing to release Kaldi-style reproducible recipes using open source and publicly available datasets for all the ASR, ST, and TTS tasks for the community to succeed our exciting outcomes.

cs.CL↗

arXiv2019

Low-PAPR Multi-channel OOK Waveform for IEEE 802.11ba Wake-up Radio

The peak-to-average-power ratio (PAPR) of the frequency domain multiplexed wake-up signals (WUSs) specified in IEEE P802.11ba can be very large and difficult to manage since it depends on the number and allocation of the active channels, and the data rate on each channel. To address this issue, we propose a transmission scheme based on complementary sequences (CSs) for multiple WUSs multiplexed in the frequency domain. We discuss how to construct CSs compatible with the framework of IEEE P802.11ba by exploiting a recursive Golay complementary pair (GCP) construction to reduce the instantaneous power fluctuations in time. We compare the proposed scheme with the other options under a non-linear power amplifier (PA) distortion. Numerical results show that the proposed scheme can lower the PAPR of the transmitted signal in frequency division multiple access (FDMA) scenarios more than 3 dB and yields a superior error rate performance under severe PA distortion.

cs.IT↗

arXiv2019

Exploring Methods for the Automatic Detection of Errors in Manual Transcription

Quality of data plays an important role in most deep learning tasks. In the speech community, transcription of speech recording is indispensable. Since the transcription is usually generated artificially, automatically finding errors in manual transcriptions not only saves time and labors but benefits the performance of tasks that need the training process. Inspired by the success of hybrid automatic speech recognition using both language model and acoustic model, two approaches of automatic error detection in the transcriptions have been explored in this work. Previous study using a biased language model approach, relying on a strong transcription-dependent language model, has been reviewed. In this work, we propose a novel acoustic model based approach, focusing on the phonetic sequence of speech. Both methods have been evaluated on a completely real dataset, which was originally transcribed with errors and strictly corrected manually afterwards.

cs.CL↗

arXiv2019

In-Edge AI: Intelligentizing Mobile Edge Computing, Caching and Communication by Federated Learning

Recently, along with the rapid development of mobile communication technology, edge computing theory and techniques have been attracting more and more attentions from global researchers and engineers, which can significantly bridge the capacity of cloud and requirement of devices by the network edges, and thus can accelerate the content deliveries and improve the quality of mobile services. In order to bring more intelligence to the edge systems, compared to traditional optimization methodology, and driven by the current deep learning techniques, we propose to integrate the Deep Reinforcement Learning techniques and Federated Learning framework with the mobile edge systems, for optimizing the mobile edge computing, caching and communication. And thus, we design the "In-Edge AI" framework in order to intelligently utilize the collaboration among devices and edge nodes to exchange the learning parameters for a better training and inference of the models, and thus to carry out dynamic system-level optimization and application-level enhancement while reducing the unnecessary system communication load. "In-Edge AI" is evaluated and proved to have near-optimal performance but relatively low overhead of learning, while the system is cognitive and adaptive to the mobile communication systems. Finally, we discuss several related challenges and opportunities for unveiling a promising upcoming future of "In-Edge AI".

cs.NI↗

arXiv2019

An Investigation of End-to-End Multichannel Speech Recognition for Reverberant and Mismatch Conditions

Sequence-to-sequence (S2S) modeling is becoming a popular paradigm for automatic speech recognition (ASR) because of its ability to jointly optimize all the conventional ASR components in an end-to-end (E2E) fashion. This report investigates the ability of E2E ASR from standard close-talk to far-field applications by encompassing entire multichannel speech enhancement and ASR components within the S2S model. There have been previous studies on jointly optimizing neural beamforming alongside E2E ASR for denoising. It is clear from both recent challenge outcomes and successful products that far-field systems would be incomplete without solving both denoising and dereverberation simultaneously. This report uses a recently developed architecture for far-field ASR by composing neural extensions of dereverberation and beamforming modules with the S2S ASR module as a single differentiable neural network and also clearly defining the role of each subnetwork. The original implementation of this architecture was successfully applied to the noisy speech recognition task (CHiME-4), while we applied this implementation to noisy reverberant tasks (DIRHA and REVERB). Our investigation shows that the method achieves better performance than conventional pipeline methods on the DIRHA English dataset and comparable performance on the REVERB dataset. It also has additional advantages of being neither iterative nor requiring parallel noisy and clean speech data.

eess.AS↗

arXiv2019

Attention Based Glaucoma Detection: A Large-scale Database and CNN Model

Recently, the attention mechanism has been successfully applied in convolutional neural networks (CNNs), significantly boosting the performance of many computer vision tasks. Unfortunately, few medical image recognition approaches incorporate the attention mechanism in the CNNs. In particular, there exists high redundancy in fundus images for glaucoma detection, such that the attention mechanism has potential in improving the performance of CNN-based glaucoma detection. This paper proposes an attention-based CNN for glaucoma detection (AG-CNN). Specifically, we first establish a large-scale attention based glaucoma (LAG) database, which includes 5,824 fundus images labeled with either positive glaucoma (2,392) or negative glaucoma (3,432). The attention maps of the ophthalmologists are also collected in LAG database through a simulated eye-tracking experiment. Then, a new structure of AG-CNN is designed, including an attention prediction subnet, a pathological area localization subnet and a glaucoma classification subnet. Different from other attention-based CNN methods, the features are also visualized as the localized pathological area, which can advance the performance of glaucoma detection. Finally, the experiment results show that the proposed AG-CNN approach significantly advances state-of-the-art glaucoma detection.

cs.CV↗

arXiv2019

Stream attention-based multi-array end-to-end speech recognition

Automatic Speech Recognition (ASR) using multiple microphone arrays has achieved great success in the far-field robustness. Taking advantage of all the information that each array shares and contributes is crucial in this task. Motivated by the advances of joint Connectionist Temporal Classification (CTC)/attention mechanism in the End-to-End (E2E) ASR, a stream attention-based multi-array framework is proposed in this work. Microphone arrays, acting as information streams, are activated by separate encoders and decoded under the instruction of both CTC and attention networks. In terms of attention, a hierarchical structure is adopted. On top of the regular attention networks, stream attention is introduced to steer the decoder toward the most informative encoders. Experiments have been conducted on AMI and DIRHA multi-array corpora using the encoder-decoder architecture. Compared with the best single-array results, the proposed framework has achieved relative Word Error Rates (WERs) reduction of 3.7% and 9.7% in the two datasets, respectively, which is better than conventional strategies as well.

cs.CL↗

arXiv2018

Multi-encoder multi-resolution framework for end-to-end speech recognition

Attention-based methods and Connectionist Temporal Classification (CTC) network have been promising research directions for end-to-end Automatic Speech Recognition (ASR). The joint CTC/Attention model has achieved great success by utilizing both architectures during multi-task training and joint decoding. In this work, we present a novel Multi-Encoder Multi-Resolution (MEMR) framework based on the joint CTC/Attention model. Two heterogeneous encoders with different architectures, temporal resolutions and separate CTC networks work in parallel to extract complimentary acoustic information. A hierarchical attention mechanism is then used to combine the encoder-level information. To demonstrate the effectiveness of the proposed model, experiments are conducted on Wall Street Journal (WSJ) and CHiME-4, resulting in relative Word Error Rate (WER) reduction of 18.0-32.1%. Moreover, the proposed MEMR model achieves 3.6% WER in the WSJ eval92 test set, which is the best WER reported for an end-to-end system on this benchmark.

cs.CL↗

arXiv2018

Cognitive-LPWAN: Towards Intelligent Wireless Services in Hybrid Low Power Wide Area Networks

The relentless development of the Internet of Things (IoT) communication technologies and the gradual maturity of Artificial Intelligence (AI) have led to a powerful cognitive computing ability. Users can now access efficient and convenient smart services in smart-city, green-IoT and heterogeneous networks. AI has been applied in various areas, including the intelligent household, advanced health-care, automatic driving and emotional interactions. This paper focuses on current wireless-communication technologies, including cellular-communication technologies (4G, 5G), low-power wide-area (LPWA) technologies with an unlicensed spectrum (LoRa, SigFox), and other LPWA technologies supported by 3GPP working with an authorized spectrum (EC-GSM, LTE-M, NB-IoT). We put forward a cognitive low-power wide-area-network (Cognitive-LPWAN) architecture to safeguard stable and efficient communications in a heterogeneous IoT. To ensure that the user can employ the AI efficiently and conveniently, we realize a variety of LPWA technologies to safeguard the network layer. In addition, to balance the demand for heterogeneous IoT devices with the communication delay and energy consumption, we put forward the AI-enabled LPWA hybrid method, starting from the perspective of traffic control. The AI algorithm provides the smart control of wireless-communication technology, intelligent applications and services for the choice of different wireless-communication technologies. As an example, we consider the AIWAC emotion interaction system, build the Cognitive-LPWAN and test the proposed AI-enabled LPWA hybrid method. The experimental results show that our scheme can meet the demands of communication-delay applications. Cognitive-LPWAN selects appropriate communication technologies to achieve a better interaction experience.

cs.NI↗

arXiv2018

A Deep Learning Approach to Denoise Optical Coherence Tomography Images of the Optic Nerve Head

Purpose: To develop a deep learning approach to de-noise optical coherence tomography (OCT) B-scans of the optic nerve head (ONH). Methods: Volume scans consisting of 97 horizontal B-scans were acquired through the center of the ONH using a commercial OCT device (Spectralis) for both eyes of 20 subjects. For each eye, single-frame (without signal averaging), and multi-frame (75x signal averaging) volume scans were obtained. A custom deep learning network was then designed and trained with 2,328 "clean B-scans" (multi-frame B-scans), and their corresponding "noisy B-scans" (clean B-scans + gaussian noise) to de-noise the single-frame B-scans. The performance of the de-noising algorithm was assessed qualitatively, and quantitatively on 1,552 B-scans using the signal to noise ratio (SNR), contrast to noise ratio (CNR), and mean structural similarity index metrics (MSSIM). Results: The proposed algorithm successfully denoised unseen single-frame OCT B-scans. The denoised B-scans were qualitatively similar to their corresponding multi-frame B-scans, with enhanced visibility of the ONH tissues. The mean SNR increased from $4.02 \pm 0.68$ dB (single-frame) to $8.14 \pm 1.03$ dB (denoised). For all the ONH tissues, the mean CNR increased from $3.50 \pm 0.56$ (single-frame) to $7.63 \pm 1.81$ (denoised). The MSSIM increased from $0.13 \pm 0.02$ (single frame) to $0.65 \pm 0.03$ (denoised) when compared with the corresponding multi-frame B-scans. Conclusions: Our deep learning algorithm can denoise a single-frame OCT B-scan of the ONH in under 20 ms, thus offering a framework to obtain superior quality OCT B-scans with reduced scanning times and minimal patient discomfort.

cs.CV↗

arXiv2018

fc: A Package for Generalized Function Composition Using Standard Evaluation

In this article, we present a new R package fc that provides a streamlined, standard evaluation-based approach to function composition. Using fc, a sequence of functions can be composed together such that returned objects from composed functions are used as intermediate values directly passed to the next function. Unlike with magrittr and purrr, no intermediate values need to be stored. When benchmarked, functions composed using fc achieve favorable runtimes in comparison to other implementations.

stat.CO↗

arXiv2018

Predictive Accuracy of Markers or Risk Scores for Interval Censored Survival Data

Methods for the evaluation of the predictive accuracy of biomarkers with respect to survival outcomes subject to right censoring have been discussed extensively in the literature. In cancer and other diseases, survival outcomes are commonly subject to interval censoring by design or due to the follow up schema. In this paper, we present an estimator for the area under the time-dependent receiver operating characteristic ROC curve for interval censored data based on a nonparametric sieve maximum likelihood approach. We establish the asymptotic properties of the proposed estimator, and illustrate its finite-sample properties using a simulation study. The application of our method is illustrated using data from a cancer clinical study. An open-source R package to implement the proposed method is available on CRAN.

stat.ME↗

arXiv2018

D2D Big Data: Content Deliveries over Wireless Device-to-Device Sharing in Large Scale Mobile Networks

Recently the topic of how to effectively offload cellular traffic onto device-to-device (D2D) sharing among users in proximity has been gaining more and more attention of global researchers and engineers. Users utilize wireless short-range D2D communications for sharing contents locally, due to not only the rapid sharing experience and free cost, but also high accuracy on deliveries of interesting and popular contents, as well as strong social impacts among friends. Nevertheless, the existing related studies are mostly confined to small-scale datasets, limited dimensions of user features, or unrealistic assumptions and hypotheses on user behaviors. In this article, driven by emerging Big Data techniques, we propose to design a big data platform, named D2D Big Data, in order to encourage the wireless D2D communications among users effectively, to promote contents for providers accurately, and to carry out offloading intelligence for operators efficiently. We deploy a big data platform and further utilize a large-scale dataset (3.56 TBytes) from a popular D2D sharing application (APP), which contains 866 million D2D sharing activities on 4.5 million files disseminated via nearly 850 million users in 13 weeks. By abstracting and analyzing multidimensional features, including online behaviors, content properties, location relations, structural characteristics, meeting dynamics, social arborescence, privacy preservation policies and so on, we verify and evaluate the D2D Big Data platform regarding predictive content propagating coverage. Finally, we discuss challenges and opportunities regarding D2D Big Data and propose to unveil a promising upcoming future of wireless D2D communications.

cs.NI↗

arXiv2017

Stream Attention for far-field multi-microphone ASR

A stream attention framework has been applied to the posterior probabilities of the deep neural network (DNN) to improve the far-field automatic speech recognition (ASR) performance in the multi-microphone configuration. The stream attention scheme has been realized through an attention vector, which is derived by predicting the ASR performance from the phoneme posterior distribution of individual microphone stream, focusing the recognizer's attention to more reliable microphones. Investigation on the various ASR performance measures has been carried out using the real recorded dataset. Experiments results show that the proposed framework has yielded substantial improvements in word error rate (WER).

cs.SD↗

arXiv2017

T-DB: Toward Fully Functional Transparent Encrypted Databases in DBaaS Framework

Individuals and organizations tend to migrate their data to clouds, especially in a DataBase as a Service (DBaaS) pattern. The major obstacle is the conflict between secrecy and utilization of the relational database to be outsourced. We address this obstacle with a Transparent DataBase (T-DB) system strictly following the unmodified DBaaS framework. A database owner outsources an encrypted database to a cloud platform, needing only to store the secret keys for encryption and an empty table header for the database; the database users can make almost all types of queries on the encrypted database as usual; and the cloud can process ciphertext queries as if the database were not encrypted. Experimentations in realistic cloud environments demonstrate that T-DB has perfect query answer precision and outstanding performance.

cs.DB↗

arXiv2017

On Accelerating the Regularized Alternating Least Square Algorithm for Tensors

In this paper, we discuss the acceleration of the regularized alternating least square (RALS) algorithm for tensor approximation. We propose a fast iterative method using a Aitken-Stefensen like updates for the regularized algorithm. Through numerical experiments, the fast algorithm demonstrate a faster convergence rate for the accelerated version in comparison to both the standard and regularized alternating least squares algorithms. In addition, we analyze the global convergence based on the Kurdyka- Lojasiewicz inequality as well as show that the RALS algorithm has a linear local convergence rate.

math.NA↗