arXiv subjects

Yu Wu

Publications and source records attributed to Yu Wu.

At least 181 records · Page 10Linked to original sources

arXiv2023

Two-Stream Network for Sign Language Recognition and Translation

Sign languages are visual languages using manual articulations and non-manual elements to convey information. For sign language recognition and translation, the majority of existing approaches directly encode RGB videos into hidden representations. RGB videos, however, are raw signals with substantial visual redundancy, leading the encoder to overlook the key information for sign language understanding. To mitigate this problem and better incorporate domain knowledge, such as handshape and body movement, we introduce a dual visual encoder containing two separate streams to model both the raw videos and the keypoint sequences generated by an off-the-shelf keypoint estimator. To make the two streams interact with each other, we explore a variety of techniques, including bidirectional lateral connection, sign pyramid network with auxiliary supervision, and frame-level self-distillation. The resulting model is called TwoStream-SLR, which is competent for sign language recognition (SLR). TwoStream-SLR is extended to a sign language translation (SLT) model, TwoStream-SLT, by simply attaching an extra translation network. Experimentally, our TwoStream-SLR and TwoStream-SLT achieve state-of-the-art performance on SLR and SLT tasks across a series of datasets including Phoenix-2014, Phoenix-2014T, and CSL-Daily. Code and models are available at: https://github.com/FangyunWei/SLRT.

cs.CV↗

arXiv2023

Speak Foreign Languages with Your Own Voice: Cross-Lingual Neural Codec Language Modeling

We propose a cross-lingual neural codec language model, VALL-E X, for cross-lingual speech synthesis. Specifically, we extend VALL-E and train a multi-lingual conditional codec language model to predict the acoustic token sequences of the target language speech by using both the source language speech and the target language text as prompts. VALL-E X inherits strong in-context learning capabilities and can be applied for zero-shot cross-lingual text-to-speech synthesis and zero-shot speech-to-speech translation tasks. Experimental results show that it can generate high-quality speech in the target language via just one speech utterance in the source language as a prompt while preserving the unseen speaker's voice, emotion, and acoustic environment. Moreover, VALL-E X effectively alleviates the foreign accent problems, which can be controlled by a language ID. Audio samples are available at \url{https://aka.ms/vallex}.

cs.CL↗

arXiv2023

Discrete Contrastive Diffusion for Cross-Modal Music and Image Generation

Diffusion probabilistic models (DPMs) have become a popular approach to conditional generation, due to their promising results and support for cross-modal synthesis. A key desideratum in conditional synthesis is to achieve high correspondence between the conditioning input and generated output. Most existing methods learn such relationships implicitly, by incorporating the prior into the variational lower bound. In this work, we take a different route -- we explicitly enhance input-output connections by maximizing their mutual information. To this end, we introduce a Conditional Discrete Contrastive Diffusion (CDCD) loss and design two contrastive diffusion mechanisms to effectively incorporate it into the denoising process, combining the diffusion training and contrastive learning for the first time by connecting it with the conventional variational objectives. We demonstrate the efficacy of our approach in evaluations with diverse multimodal conditional synthesis tasks: dance-to-music generation, text-to-image synthesis, as well as class-conditioned image synthesis. On each, we enhance the input-output correspondence and achieve higher or competitive general synthesis quality. Furthermore, the proposed approach improves the convergence of diffusion models, reducing the number of required diffusion steps by more than 35% on two benchmarks, significantly increasing the inference speed.

cs.CV↗

arXiv2023

RF Injection Locking of THz Metasurface Quantum-Cascade VECSEL

RF injection locking and spectral broadening of a terahertz (THz) quantum-cascade vertical-external-cavity surface-emitting laser (QC-VECSEL) is demonstrated. An intra-cryostat VECSEL focusing cavity design is used to enable continuous-wave lasing with a cavity length over 30 mm which corresponds to a round-trip frequency near 5 GHz. Strong RF current modulation is injected to the QC-metasurface electrical bias to pull and lock the round-trip frequency. The injection locking range at various RF injection powers is recorded and compared with the injection locking theory. Moreover, the lasing spectrum broadens from 14 GHz in free-running mode to a maximum spectral width around 100 GHz with 20 dBm of injected RF power. This experimental setup is suitable for further exploration of active mode-locking and picosecond pulse generation in THz QC-VECSELs.

physics.optics↗

arXiv2023

Neural Codec Language Models are Zero-Shot Text to Speech Synthesizers

We introduce a language modeling approach for text to speech synthesis (TTS). Specifically, we train a neural codec language model (called Vall-E) using discrete codes derived from an off-the-shelf neural audio codec model, and regard TTS as a conditional language modeling task rather than continuous signal regression as in previous work. During the pre-training stage, we scale up the TTS training data to 60K hours of English speech which is hundreds of times larger than existing systems. Vall-E emerges in-context learning capabilities and can be used to synthesize high-quality personalized speech with only a 3-second enrolled recording of an unseen speaker as an acoustic prompt. Experiment results show that Vall-E significantly outperforms the state-of-the-art zero-shot TTS system in terms of speech naturalness and speaker similarity. In addition, we find Vall-E could preserve the speaker's emotion and acoustic environment of the acoustic prompt in synthesis. See https://aka.ms/valle for demos of our work.

cs.CL↗

arXiv2022

Generative Graph Neural Networks for Link Prediction

Inferring missing links or detecting spurious ones based on observed graphs, known as link prediction, is a long-standing challenge in graph data analysis. With the recent advances in deep learning, graph neural networks have been used for link prediction and have achieved state-of-the-art performance. Nevertheless, existing methods developed for this purpose are typically discriminative, computing features of local subgraphs around two neighboring nodes and predicting potential links between them from the perspective of subgraph classification. In this formalism, the selection of enclosing subgraphs and heuristic structural features for subgraph classification significantly affects the performance of the methods. To overcome this limitation, this paper proposes a novel and radically different link prediction algorithm based on the network reconstruction theory, called GraphLP. Instead of sampling positive and negative links and heuristically computing the features of their enclosing subgraphs, GraphLP utilizes the feature learning ability of deep-learning models to automatically extract the structural patterns of graphs for link prediction under the assumption that real-world graphs are not locally isolated. Moreover, GraphLP explores high-order connectivity patterns to utilize the hierarchical organizational structures of graphs for link prediction. Our experimental results on all common benchmark datasets from different applications demonstrate that the proposed method consistently outperforms other state-of-the-art methods. Unlike the discriminative neural network models used for link prediction, GraphLP is generative, which provides a new paradigm for neural-network-based link prediction.

cs.SI↗

arXiv2022

How to Share: Balancing Layer and Chain Sharing in Industrial Microservice Deployment

With the rapid development of smart manufacturing, edge computing-oriented microservice platforms are emerging as an important part of production control. In the containerized deployment of microservices, layer sharing can reduce the huge bandwidth consumption caused by image pulling, and chain sharing can reduce communication overhead caused by communication between microservices. The two sharing methods use the characteristics of each microservice to share resources during deployment. However, due to the limited resources of edge servers, it is difficult to meet the optimization goals of the two methods at the same time. Therefore, it is of critical importance to realize the improvement of service response efficiency by balancing the two sharing methods. This paper studies the optimal microservice deployment strategy that can balance layer sharing and chain sharing of microservices. We build a problem that minimizes microservice image pull delay and communication overhead and transform the problem into a linearly constrained integer quadratic programming problem through model reconstruction. A deployment strategy is obtained through the successive convex approximation (SCA) method. Experimental results show that the proposed deployment strategy can balance the two resource sharing methods. When the two sharing methods are equally considered, the average image pull delay can be reduced to 65% of the baseline, and the average communication overhead can be reduced to 30% of the baseline.

eess.SY↗

arXiv2022

BEATs: Audio Pre-Training with Acoustic Tokenizers

The massive growth of self-supervised learning (SSL) has been witnessed in language, vision, speech, and audio domains over the past few years. While discrete label prediction is widely adopted for other modalities, the state-of-the-art audio SSL models still employ reconstruction loss for pre-training. Compared with reconstruction loss, semantic-rich discrete label prediction encourages the SSL model to abstract the high-level audio semantics and discard the redundant details as in human perception. However, a semantic-rich acoustic tokenizer for general audio pre-training is usually not straightforward to obtain, due to the continuous property of audio and unavailable phoneme sequences like speech. To tackle this challenge, we propose BEATs, an iterative audio pre-training framework to learn Bidirectional Encoder representation from Audio Transformers, where an acoustic tokenizer and an audio SSL model are optimized by iterations. In the first iteration, we use random projection as the acoustic tokenizer to train an audio SSL model in a mask and label prediction manner. Then, we train an acoustic tokenizer for the next iteration by distilling the semantic knowledge from the pre-trained or fine-tuned audio SSL model. The iteration is repeated with the hope of mutual promotion of the acoustic tokenizer and audio SSL model. The experimental results demonstrate our acoustic tokenizers can generate discrete labels with rich audio semantics and our audio SSL models achieve state-of-the-art results across various audio classification benchmarks, even outperforming previous models that use more training data and model parameters significantly. Specifically, we set a new state-of-the-art mAP 50.6% on AudioSet-2M for audio-only models without using any external data, and 98.1% accuracy on ESC-50. The code and pre-trained models are available at https://aka.ms/beats.

eess.AS↗

arXiv2022

Artificial Intelligence Security Competition (AISC)

The security of artificial intelligence (AI) is an important research area towards safe, reliable, and trustworthy AI systems. To accelerate the research on AI security, the Artificial Intelligence Security Competition (AISC) was organized by the Zhongguancun Laboratory, China Industrial Control Systems Cyber Emergency Response Team, Institute for Artificial Intelligence, Tsinghua University, and RealAI as part of the Zhongguancun International Frontier Technology Innovation Competition (https://www.zgc-aisc.com/en). The competition consists of three tracks, including Deepfake Security Competition, Autonomous Driving Security Competition, and Face Recognition Security Competition. This report will introduce the competition rules of these three tracks and the solutions of top-ranking teams in each track.

cs.CR↗

arXiv2022

Speech separation with large-scale self-supervised learning

Self-supervised learning (SSL) methods such as WavLM have shown promising speech separation (SS) results in small-scale simulation-based experiments. In this work, we extend the exploration of the SSL-based SS by massively scaling up both the pre-training data (more than 300K hours) and fine-tuning data (10K hours). We also investigate various techniques to efficiently integrate the pre-trained model with the SS network under a limited computation budget, including a low frame rate SSL model training setup and a fine-tuning scheme using only the part of the pre-trained model. Compared with a supervised baseline and the WavLM-based SS model using feature embeddings obtained with the previously released 94K hours trained WavLM, our proposed model obtains 15.9% and 11.2% of relative word error rate (WER) reductions, respectively, for a simulated far-field speech mixture test set. For conversation transcription on real meeting recordings using continuous speech separation, the proposed model achieves 6.8% and 10.6% of relative WER reductions over the purely supervised baseline on AMI and ICSI evaluation sets, respectively, while reducing the computational cost by 38%.

eess.AS↗

arXiv2022

Turning Silver into Gold: Domain Adaptation with Noisy Labels for Wearable Cardio-Respiratory Fitness Prediction

Deep learning models have shown great promise in various healthcare applications. However, most models are developed and validated on small-scale datasets, as collecting high-quality (gold-standard) labels for health applications is often costly and time-consuming. As a result, these models may suffer from overfitting and not generalize well to unseen data. At the same time, an extensive amount of data with imprecise labels (silver-standard) is starting to be generally available, as collected from inexpensive wearables like accelerometers and electrocardiography sensors. These currently underutilized datasets and labels can be leveraged to produce more accurate clinical models. In this work, we propose UDAMA, a novel model with two key components: Unsupervised Domain Adaptation and Multi-discriminator Adversarial training, which leverage noisy data from source domain (the silver-standard dataset) to improve gold-standard modeling. We validate our framework on the challenging task of predicting lab-measured maximal oxygen consumption (VO$_{2}$max), the benchmark metric of cardio-respiratory fitness, using free-living wearable sensor data from two cohort studies as inputs. Our experiments show that the proposed framework achieves the best performance of corr = 0.665 $\pm$ 0.04, paving the way for accurate fitness estimation at scale.

eess.SP↗

arXiv2022

Exploring WavLM on Speech Enhancement

There is a surge in interest in self-supervised learning approaches for end-to-end speech encoding in recent years as they have achieved great success. Especially, WavLM showed state-of-the-art performance on various speech processing tasks. To better understand the efficacy of self-supervised learning models for speech enhancement, in this work, we design and conduct a series of experiments with three resource conditions by combining WavLM and two high-quality speech enhancement systems. Also, we propose a regression-based WavLM training objective and a noise-mixing data configuration to further boost the downstream enhancement performance. The experiments on the DNS challenge dataset and a simulation dataset show that the WavLM benefits the speech enhancement task in terms of both speech quality and speech recognition accuracy, especially for low fine-tuning resources. For the high fine-tuning resource condition, only the word error rate is substantially improved.

eess.AS↗

arXiv2022

LongFNT: Long-form Speech Recognition with Factorized Neural Transducer

Traditional automatic speech recognition~(ASR) systems usually focus on individual utterances, without considering long-form speech with useful historical information, which is more practical in real scenarios. Simply attending longer transcription history for a vanilla neural transducer model shows no much gain in our preliminary experiments, since the prediction network is not a pure language model. This motivates us to leverage the factorized neural transducer structure, containing a real language model, the vocabulary predictor. We propose the {LongFNT-Text} architecture, which fuses the sentence-level long-form features directly with the output of the vocabulary predictor and then embeds token-level long-form features inside the vocabulary predictor, with a pre-trained contextual encoder RoBERTa to further boost the performance. Moreover, we propose the {LongFNT} architecture by extending the long-form speech to the original speech input and achieve the best performance. The effectiveness of our LongFNT approach is validated on LibriSpeech and GigaSpeech corpora with 19% and 12% relative word error rate~(WER) reduction, respectively.

cs.SD↗

arXiv2022

Investigation of the Pauli paramagnetic effect in systematically tuned NbN thin films

Superconductivity and the normal-state properties of NbN films can be tuned in a wide range, supplying a suitable platform to investigate the systematical evolution of the superconducting performances. Herein, we report the upper critical field of NbN films in both the vertical ($B\perp$ film) and parallel ($B\parallel$ film) orientations over a wide temperature range. Eight samples with the superconducting critical temperature $T_c$ ranging from 2.5 K to 9.8 K are studied. Meanwhile, the normal-state resistivity is tuned by more than six times by changing the conditions of the film growth. It is found that the magnitudes of the upper critical field in both field directions ($B_{c2}^{\perp}$ and $B_{c2}^{\parallel}$) exceed the paramagnetic limiting field $B_p$. The temperature dependent $B_{c2}^{\perp}$ can be described by the extended Werthamer--Helfand--Hohenberg (WHH) model considering the Pauli spin paramagnetism. Meanwhile, the $B_{c2}^{\parallel}$-$T$ data shows the feature of two-dimensional superconductivity in the temperature range near $T_c$. The evolution of the obtained Maki parameter with other parameters, such as the slope of the upper critical field near $T_c$ and the normal-state resistivity, are discussed to reveal the characteristics of the Pauli paramagnetic effect in this system.

cond-mat.supr-con↗

arXiv2022

Robust Quantum Griffiths Singularity at above 1.5 Kelvin in Nitride Thin Films

Quantum Griffiths singularity (QGS), which is closely correlated with the quenched disorder, is characterized by the divergence of the dynamical critical exponent and the presence of activated scaling behavior. Typically such a quantum phenomenon is rather rare and only observed in extremely low temperatures. Here we report the experimental observation of a robust QGS in nitride thin films, NbN, which survives in a rather high temperature range. The electrical transport propertied were measured under the magnetic field up to 12 T. The field induced superconductor-metal transitions were observed with the continuously changed transition points with the decrease of temperature. The dynamical critical exponent based on the conventional power-law scaling reveals a divergent trend when approaching the the low temperature limit. Moreover, the temperature and field dependence of sheet resistance can be described by the activated scaling analysis in the temperature region up to 1.6 K $\leq T \leq$ 4.0 K. We argue that the robustness of QGS in the present system originates from the Pauli paramagnetic effect due to the strong coupling between the magnetic field and the spin degree of freedom.

cond-mat.supr-con↗

arXiv2022

Real-time Speech Interruption Analysis: From Cloud to Client Deployment

Meetings are an essential form of communication for all types of organizations, and remote collaboration systems have been much more widely used since the COVID-19 pandemic. One major issue with remote meetings is that it is challenging for remote participants to interrupt and speak. We have recently developed the first speech interruption analysis model, which detects failed speech interruptions, shows very promising performance, and is being deployed in the cloud. To deliver this feature in a more cost-efficient and environment-friendly way, we reduced the model complexity and size to ship the WavLM_SI model in client devices. In this paper, we first describe how we successfully improved the True Positive Rate (TPR) at a 1% False Positive Rate (FPR) from 50.9% to 68.3% for the failed speech interruption detection model by training on a larger dataset and fine-tuning. We then shrank the model size from 222.7 MB to 9.3 MB with an acceptable loss in accuracy and reduced the complexity from 31.2 GMACS (Giga Multiply-Accumulate Operations per Second) to 4.3 GMACS. We also estimated the environmental impact of the complexity reduction, which can be used as a general guideline for large Transformer-based models, and thus make those models more accessible with less computation overhead.

cs.CL↗

arXiv2022

Longitudinal cardio-respiratory fitness prediction through wearables in free-living environments

Cardiorespiratory fitness is an established predictor of metabolic disease and mortality. Fitness is directly measured as maximal oxygen consumption (VO$_{2}max$), or indirectly assessed using heart rate responses to standard exercise tests. However, such testing is costly and burdensome because it requires specialized equipment such as treadmills and oxygen masks, limiting its utility. Modern wearables capture dynamic real-world data which could improve fitness prediction. In this work, we design algorithms and models that convert raw wearable sensor data into cardiorespiratory fitness estimates. We validate these estimates' ability to capture fitness profiles in free-living conditions using the Fenland Study (N=11,059), along with its longitudinal cohort (N=2,675), and a third external cohort using the UK Biobank Validation Study (N=181) who underwent maximal VO$_{2}max$ testing, the gold standard measurement of fitness. Our results show that the combination of wearables and other biomarkers as inputs to neural networks yields a strong correlation to ground truth in a holdout sample (r = 0.82, 95CI 0.80-0.83), outperforming other approaches and models and detects fitness change over time (e.g., after 7 years). We also show how the model's latent space can be used for fitness-aware patient subtyping paving the way to scalable interventions and personalized trial recruitment. These results demonstrate the value of wearables for fitness estimation that today can be measured only with laboratory tests.

cs.LG↗

arXiv2022

Foundation Transformers

A big convergence of model architectures across language, vision, speech, and multimodal is emerging. However, under the same name "Transformers", the above areas use different implementations for better performance, e.g., Post-LayerNorm for BERT, and Pre-LayerNorm for GPT and vision Transformers. We call for the development of Foundation Transformer for true general-purpose modeling, which serves as a go-to architecture for various tasks and modalities with guaranteed training stability. In this work, we introduce a Transformer variant, named Magneto, to fulfill the goal. Specifically, we propose Sub-LayerNorm for good expressivity, and the initialization strategy theoretically derived from DeepNet for stable scaling up. Extensive experiments demonstrate its superior performance and better stability than the de facto Transformer variants designed for various applications, including language modeling (i.e., BERT, and GPT), machine translation, vision pretraining (i.e., BEiT), speech recognition, and multimodal pretraining (i.e., BEiT-3).

cs.LG↗