arXiv subjects

Xiaoyu Yang

Publications and source records attributed to Xiaoyu Yang.

At least 73 records · Page 4Linked to original sources

arXiv2024

LibriheavyMix: A 20,000-Hour Dataset for Single-Channel Reverberant Multi-Talker Speech Separation, ASR and Speaker Diarization

The evolving speech processing landscape is increasingly focused on complex scenarios like meetings or cocktail parties with multiple simultaneous speakers and far-field conditions. Existing methodologies for addressing these challenges fall into two categories: multi-channel and single-channel solutions. Single-channel approaches, notable for their generality and convenience, do not require specific information about microphone arrays. This paper presents a large-scale far-field overlapping speech dataset, crafted to advance research in speech separation, recognition, and speaker diarization. This dataset is a critical resource for decoding ``Who said What and When'' in multi-talker, reverberant environments, a daunting challenge in the field. Additionally, we introduce a pipeline system encompassing speech separation, recognition, and diarization as a foundational benchmark. Evaluations on the WHAMR! dataset validate the broad applicability of the proposed data.

cs.SD↗

arXiv2024

SUBLLM: A Novel Efficient Architecture with Token Sequence Subsampling for LLM

While Large Language Models (LLMs) have achieved remarkable success in various fields, the efficiency of training and inference remains a major challenge. To address this issue, we propose SUBLLM, short for Subsampling-Upsampling-Bypass Large Language Model, an innovative architecture that extends the core decoder-only framework by incorporating subsampling, upsampling, and bypass modules. The subsampling modules are responsible for shortening the sequence, while the upsampling modules restore the sequence length, and the bypass modules enhance convergence. In comparison to LLaMA, the proposed SUBLLM exhibits significant enhancements in both training and inference speeds as well as memory usage, while maintaining competitive few-shot performance. During training, SUBLLM increases speeds by 26% and cuts memory by 10GB per GPU. In inference, it boosts speeds by up to 37% and reduces memory by 1GB per GPU. The training and inference speeds can be enhanced by 34% and 52% respectively when the context window is expanded to 8192. Our code is available at https://github.com/XiaoMi/subllm.

cs.CL↗

arXiv2024

Interference Management in MIMO-ISAC Systems: A Transceiver Design Approach

The integrated sensing and communication (ISAC) system under multi-input multi-output (MIMO) architecture achieves dual functionalities of sensing and communication on the same platform by utilizing spatial gain, which provides a feasible paradigm facing spectrum congestion. However, the dual functionalities of sensing and communication operating simultaneously in the same platform bring severe interference in the ISAC systems. Facing this challenge, we propose a joint optimization framework for transmit beamforming and receive filter design for ISAC systems with MIMO architecture. We aim to maximize the signal-to-clutter-plus-noise ratio (SCNR) at the receiver while considering various constraints such as waveform similarity, power budget, and communication performance requirements to ensure the integration of the dual functionalities. In particular, the overall transmit beamforming is refined into sensing beamforming and communication beamforming, and a quadratic transformation (QT) is introduced to relax and convert the complex non-convex optimization objective. An efficient algorithm based on covariance matrix tapers (CMT) is proposed to restructure the clutter covariance matrix considering the mismatched steering vector, thereby improving the robustness of the ISAC transceiver design. Numerical simulations are provided to demonstrate the effectiveness of the proposed algorithm.

eess.SP↗

arXiv2024

Enhancing Visual Grounding and Generalization: A Multi-Task Cycle Training Approach for Vision-Language Models

Visual grounding (VG) occupies a pivotal position in multi-modality vision-language models. In this study, we propose ViLaM, a large multi-modality model, that supports multi-tasks of VG using the cycle training strategy, with abundant interaction instructions. The cycle training between referring expression generation (REG) and referring expression comprehension (REC) is introduced. It enhances the consistency between visual location and referring expressions, and addresses the need for high-quality, multi-tasks VG datasets. Moreover, multi-tasks of VG are promoted in our model, contributed by the cycle training strategy. The multi-tasks in REC encompass a range of granularities, from region-level to pixel-level, which include referring bbox detection, referring keypoints detection, and referring image segmentation. In REG, referring region classification determines the fine-grained category of the target, while referring region captioning generates a comprehensive description. Meanwhile, all tasks participate in the joint training, synergistically enhancing one another and collectively improving the overall performance of the model. Furthermore, leveraging the capabilities of large language models, ViLaM extends a wide range of instructions, thereby significantly enhancing its generalization and interaction potentials. Extensive public datasets corroborate the superior capabilities of our model in VG with muti-tasks. Additionally, validating its robust generalization, ViLaM is validated under open-set and few-shot scenarios. Especially in the medical field, our model demonstrates cross-domain robust generalization capabilities. Furthermore, we contribute a VG dataset, especially with multi-tasks. To support and encourage the community focused on VG, we have made both the dataset and our code public: https://github.com/AnonymGiant/ViLaM.

cs.CV↗

arXiv2024

Zipformer: A faster and better encoder for automatic speech recognition

The Conformer has become the most popular encoder model for automatic speech recognition (ASR). It adds convolution modules to a transformer to learn both local and global dependencies. In this work we describe a faster, more memory-efficient, and better-performing transformer, called Zipformer. Modeling changes include: 1) a U-Net-like encoder structure where middle stacks operate at lower frame rates; 2) reorganized block structure with more modules, within which we re-use attention weights for efficiency; 3) a modified form of LayerNorm called BiasNorm allows us to retain some length information; 4) new activation functions SwooshR and SwooshL work better than Swish. We also propose a new optimizer, called ScaledAdam, which scales the update by each tensor's current scale to keep the relative change about the same, and also explictly learns the parameter scale. It achieves faster convergence and better performance than Adam. Extensive experiments on LibriSpeech, Aishell-1, and WenetSpeech datasets demonstrate the effectiveness of our proposed Zipformer over other state-of-the-art ASR models. Our code is publicly available at https://github.com/k2-fsa/icefall.

eess.AS↗

arXiv2024

Developments and applications of the OPTIMADE API for materials discovery, design, and data exchange

The Open Databases Integration for Materials Design (OPTIMADE) application programming interface (API) empowers users with holistic access to a growing federation of databases, enhancing the accessibility and discoverability of materials and chemical data. Since the first release of the OPTIMADE specification (v1.0), the API has undergone significant development, leading to the upcoming v1.2 release, and has underpinned multiple scientific studies. In this work, we highlight the latest features of the API format, accompanying software tools, and provide an update on the implementation of OPTIMADE in contributing materials databases. We end by providing several use cases that demonstrate the utility of the OPTIMADE API in materials research that continue to drive its ongoing development.

cond-mat.mtrl-sci↗

arXiv2024

Moderate deviations for rough differential equations

Small noise problems are quite important for all types of stochastic differential equations. In this paper we focus on rough differential equations driven by scaled fractional Brownian rough path with Hurst parameter H between 1/4 and 1/2. We prove a moderate deviation principle for this equation as the scale parameter tends to zero.

math.PR↗

arXiv2024

Convergence of martingale solutions to the hybrid slow-fast system

This paper is devoted to studying the weak convergence for a slow-fast system with jumps modulated by Markovian switching regimes with the martingale method. However, due to the coexistence of fast component and Markovian switching regimes, the martingale method and perturbed test functions can not be applied directly. In this situation, a combination of perturbed test functions and the time discretization is applied efficiently. And the choice of appropriate perturbed test functions, which are related to the averaged coefficients, plays a decisive role. Our results also cover the case of slow-fast system without Markovian switching regimes. Finally, some examples are presented,and numerical simulations are carried out to observe a good agreement.

math.DS↗

arXiv2024

The central limit theorem for slow-fast systems with Lévy noise

We consider a slow-fast stochastic differential system with Lévy noise. We will employ the perturbed test function method to study the normal deviation of the slow-fast system. Our main result states that the deviation can be approximated by a Gaussian process and the central limit theorem is obtained for the system.

math.PR↗

arXiv2024

Moderate deviations for two-time scale systems with mixed fractional Brownian motion

This work focuses on moderate deviations for two-time scale systems with mixed fractional Brownian motion. Our proof uses the weak convergence method which is based on the variational representation formula for mixed fractional Brownian motion. Throughout this paper, Hurst parameter of fractional Brownian motion is larger than $1/2$ and the integral along the fractional Brownian motion is understood as the generalized Riemann-Stieltjes integral. First, we consider single-time scale systems with fractional Brownian motion. The key of our proof is showing the weak convergence of the controlled system. Next, we extend our method to show moderate deviations for two-time scale systems. To this goal, we combine the Khasminskii-type averaging principle and the weak convergence approach.

math.DS↗

arXiv2024

PromptASR for contextualized ASR with controllable style

Prompts are crucial to large language models as they provide context information such as topic or logical relationships. Inspired by this, we propose PromptASR, a framework that integrates prompts in end-to-end automatic speech recognition (E2E ASR) systems to achieve contextualized ASR with controllable style of transcriptions. Specifically, a dedicated text encoder encodes the text prompts and the encodings are injected into the speech encoder by cross-attending the features from two modalities. When using the ground truth text from preceding utterances as content prompt, the proposed system achieves 21.9% and 6.8% relative word error rate reductions on a book reading dataset and an in-house dataset compared to a baseline ASR system. The system can also take word-level biasing lists as prompt to improve recognition accuracy on rare words. An additional style prompt can be given to the text encoder and guide the ASR system to output different styles of transcriptions. The code is available at icefall.

eess.AS↗

arXiv2024

Libriheavy: a 50,000 hours ASR corpus with punctuation casing and context

In this paper, we introduce Libriheavy, a large-scale ASR corpus consisting of 50,000 hours of read English speech derived from LibriVox. To the best of our knowledge, Libriheavy is the largest freely-available corpus of speech with supervisions. Different from other open-sourced datasets that only provide normalized transcriptions, Libriheavy contains richer information such as punctuation, casing and text context, which brings more flexibility for system building. Specifically, we propose a general and efficient pipeline to locate, align and segment the audios in previously published Librilight to its corresponding texts. The same as Librilight, Libriheavy also has three training subsets small, medium, large of the sizes 500h, 5000h, 50000h respectively. We also extract the dev and test evaluation sets from the aligned audios and guarantee there is no overlapping speakers and books in training sets. Baseline systems are built on the popular CTC-Attention and transducer models. Additionally, we open-source our dataset creatation pipeline which can also be used to other audio alignment tasks.

eess.AS↗

arXiv2024

Large deviation principle for a two-time-scale McKean-Vlasov model with jumps

This work focus on the large deviation principle for a two-time scale McKean-Vlasov system with jumps. Based on the variational framework of the McKean-Vlasov system with jumps, it is turned into weak convergence for the controlled system. Unlike general two-time scale system, the controlled McKean-Vlasov system is related to the law of the original system, which causes difficulties in qualitative analysis. In solving this problem, employing asymptotics of the original system and a Khasminskii-type averaging principle together is efficient. Finally, it is shown that the limit is related to the Dirac measure of the solution to the ordinary differential equation.

math.PR↗

arXiv2024

Large deviation principle for slow-fast rough differential equations via controlled rough paths

We prove a large deviation principle for the slow-fast rough differential equations under the controlled rough path framework. The driver rough paths are lifted from the mixed fractional Brownian motion with Hurst parameter $H\in (1/3,1/2)$. Our approach is based on the continuity of the solution mapping and the variational framework for mixed fractional Brownian motion. By utilizing the variational representation, our problem is transformed into a qualitative property of the controlled system. In particular, the fast rough differential equation coincides with Itô SDE almost surely, which possesses a unique invariant probability measure with frozen slow component. We then demonstrate the weak convergence of the controlled slow component by averaging with respect to the invariant measure of the fast equation and exploiting the continuity of the solution mapping.

math.PR↗

arXiv2023

Mutual Information Metrics for Uplink MIMO-OFDM Integrated Sensing and Communication System

As the uplink sensing has the advantage of easy implementation, it attracts great attention in integrated sensing and communication (ISAC) system. This paper presents an uplink ISAC system based on multi-input multi-output orthogonal frequency division multiplexing (MIMO-OFDM) technology. The mutual information (MI) is introduced as a unified metric to evaluate the performance of communication and sensing. In this paper, firstly, the upper and lower bounds of communication and sensing MI are derived in details based on the interaction between communication and sensing. And the ISAC waveform is optimized by maximizing the weighted sum of sensing and communication MI. The Monte Carlo simulation results show that, compared with other waveform optimization schemes, the proposed ISAC scheme has the best overall performance.

cs.IT↗

arXiv2023

Addressing the Accuracy-Cost Tradeoff in Material Property Prediction: A Teacher-Student Strategy

Deep learning has revolutionized the process of new material discovery, with state-of-the-art models now able to predict material properties based solely on chemical compositions, thus eliminating the necessity for material structures. However, this cost-effective method has led to a trade-off in model accuracy. Specifically, the accuracy of Chemical Composition-based Property Prediction Models (CPMs) significantly lags behind that of Structure-based Property Prediction Models (SPMs). To tackle this challenge, we propose an innovative Teacher-Student (T-S) strategy, where a pre-trained SPM serves as the 'teacher' to enhance the accuracy of the CPM. Leveraging the T-S strategy, T-S CrabNet has risen to become the most accurate model among current CPMs. Initially, we demonstrated the universality of this strategy. On the Materials Project (MP) and Jarvis datasets, we validated the effectiveness of the T-S strategy in boosting the accuracy of CPMs with two distinct network structures, namely CrabNet and Roost. This led to CrabNet, under the guidance of the T-S strategy, emerging as the most accurate model among the current CPMs. Moreover, this strategy shows remarkable efficacy in small datasets. When predicting the formation energy on a small MP dataset comprising merely 5% of the samples, the T-S strategy boosted CrabNet's accuracy by 37.1%, exceeding the enhancement effect of the T-S strategy on the whole dataset.

cond-mat.mtrl-sci↗

arXiv2023

Symbol-level Integrated Sensing and Communication enabled Multiple Base Stations Cooperative Sensing

With the support of integrated sensing and communication (ISAC) technology, mobile communication system will integrate the function of wireless sensing, thereby facilitating new intelligent applications such as smart city and intelligent transportation. Due to the limited sensing accuracy and sensing range of single base station (BS), multi-BS cooperative sensing can be applied to realize high-accurate, long-range and continuous sensing, exploiting the specific advantages of large-scale networked mobile communication system. This paper proposes a cooperative sensing method suitable to mobile communication systems, which applies symbol-level sensing information fusion to estimate the location and velocity of target. With the demodulation symbols obtained from the echo signals of multiple BSs, the phase features contained in the demodulation symbols are used in the fusion procedure, which realizes cooperative sensing with the synchronization level of mobile communication system. Compared with the signal-level fusion in the area of distributed aperture coherence-synthetic radars, the requirement of synchronization is much lower. When signal-to-noise ratio (SNR) is -5 dB, it is evaluated that symbol-level multi-BS cooperative sensing effectively improves the accuracy of distance and velocity estimation of target. Compared with single-BS sensing, the accuracy of distance and velocity estimation is improved by 40% and 72%, respectively. Compared with data-level multi-BS cooperative sensing based on maximum likelihood (ML) estimation, the accuracy of location and velocity estimation is improved by 12% and 63%, respectively. This work may provide a guideline for the design of multi-BS cooperative sensing system to exploit the widely deployed networked mobile communication system.

eess.SP↗

arXiv2023

Delay-penalized CTC implemented based on Finite State Transducer

Connectionist Temporal Classification (CTC) suffers from the latency problem when applied to streaming models. We argue that in CTC lattice, the alignments that can access more future context are preferred during training, thereby leading to higher symbol delay. In this work we propose the delay-penalized CTC which is augmented with latency penalty regularization. We devise a flexible and efficient implementation based on the differentiable Finite State Transducer (FST). Specifically, by attaching a binary attribute to CTC topology, we can locate the frames that firstly emit non-blank tokens on the resulting CTC lattice, and add the frame offsets to the log-probabilities. Experimental results demonstrate the effectiveness of our proposed delay-penalized CTC, which is able to balance the delay-accuracy trade-off. Furthermore, combining the delay-penalized transducer enables the CTC model to achieve better performance and lower latency. Our work is open-sourced and publicly available https://github.com/k2-fsa/k2.

eess.AS↗