检索结果-内蒙古大学图书馆

EchoLoc: Accurate Device-Free Hand Localization Using COTS Devices

学校读者我要写书评

暂无评论

EchoLoc: Accurate Device-Free Hand Localization Using COTS D...

International Conference on Parallel processing (ICPP)

作者： Huijie Chen Fan Li Yu Wang School of Computer Science Beijing Institute of Technology Beijing Engineering Research Center for High Volume Language Information Processing and Cloud Computing Applications Beijing China Department of Computer Science College of Computing and Informatics University of North Carolina at Charlotte Charlotte NC USA

ISBN: (纸本)9781509028245

Hand tracking systems are becoming increasingly popular as a fundamental HCI approach. The trajectory of moving hand can be estimated through smoothing the position coordinates collected from continuous localization. Therefore, hand localization is a key component of any hand tracking systems. This paper presents EchoLoc, which locates the human hand by leveraging the speaker array in Commercial Off-The-Shelf (COTS) devices (i.e., a smart phone plugged with a stereo speaker). EchoLoc measures the distance from the hand to the speaker array via the Time Of Flight (TOF) of the chirp. The speaker array and hand yield a unique triangle, therefore, the hand can be localized with triangular geometry. We prototype EchoLoc on iOS as an application, and find it is capable of localization with the average resolution within five centimeters of 73% and three centimeters of 48%.

关键词： Parallel processing

Comprehensive study: How the context information of different granularity affects dialogue state tracking?

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Yang, Puhai Huang, Heyan Mao, Xian-Ling School of Computer Science and Technology Beijing Institute of Technology Beijing China Beijing Engineering Research Center of High Volume Language Information Processing and Cloud Computing Applications Beijing China Southeast Academy of Information Technology Beijing Institute of Technology Fujian China

Dialogue state tracking (DST) plays a key role in task-oriented dialogue systems to monitor the user's goal. In general, there are two strategies to track a dialogue state: predicting it from scratch and updating it from previous state. The scratch-based strategy obtains each slot value by inquiring all the dialogue history, and the previous-based strategy relies on the current turn dialogue to update the previous dialogue state. However, it is hard for the scratch-based strategy to correctly track short-dependency dialogue state because of noise;meanwhile, the previous-based strategy is not very useful for long-dependency dialogue state tracking. Obviously, it plays different roles for the context information of different granularity to track different kinds of dialogue states. Thus, in this paper, we will study and discuss how the context information of different granularity affects dialogue state tracking. First, we explore how greatly different granularities affect dialogue state tracking. Then, we further discuss how to combine multiple granularities for dialogue state tracking. Finally, we apply the findings about context granularity to few-shot learning scenario. Besides, we have publicly released all codes. Copyright © 2021, The Authors. All rights reserved.

关键词： Speech processing

MindLLM: Lightweight Large language Model Pre-Training, Evaluation and Domain application

学校读者我要写书评

暂无评论

SSRN 2024年

作者： Yang, Yizhe Sun, Huashan Li, Jiawei Liu, Runheng Li, Yinghao Liu, Yuhang Gao, Yang Huang, Heyan School of Computer Science Beijing Institute of Technology Beijing China Beijing Engineering Research Center of High Volume Language Information Processing and Cloud Computing Applications Beijing China Beijing Institute of Technology Southeast Academy of Information Technology Fujian Putian China

Large language Models (LLMs) have demonstrated remarkable performance across various natural language tasks, marking significant strides towards general artificial intelligence. While general artificial intelligence is leveraged by developing increasingly large-scale models, there could be another branch to develop lightweight custom models that better serve certain domains, taking into account the high cost of training and deploying LLMs and the scarcity of resources. In this paper, we present MindLLM, a novel series of bilingual lightweight large language models, trained from scratch, alleviating such burdens by offering models with 1.3 billion and 3 billion parameters. A thorough account of experiences accrued during large model development is given, covering every step of the process, including data construction, model architecture, evaluation, and applications. Such insights are hopefully valuable for fellow academics and developers. MindLLM consistently matches or surpasses the performance of other open-source larger models on some public benchmarks. We also introduce an innovative instruction tuning framework tailored for smaller models to enhance their capabilities efficiently. Moreover, we explore the application of MindLLM in specific vertical domains such as law and finance, underscoring the agility and adaptability of our lightweight models. © 2024, The Authors. All rights reserved.

关键词： Artificial intelligence

Sle Diagnosis Research Based on Sers Combined with a Multi-Modal Fusion Method

学校读者我要写书评

暂无评论

SSRN 2023年

作者： Huang, Yuhao Chen, Chen Chang, Chenjie Cheng, Zhiyuan Liu, Yang Chen, Cheng Lv, Xiaoyi College of Software Xinjiang University Xinjiang Urumqi830046 China College of Information Science and Engineering Xinjiang University Urumqi830046 China Key Laboratory of Signal Detection and Processing Xinjiang University Urumqi830046 China Xinjiang Cloud Computing Application Laboratory Xinjiang Cloud Computing Engineering Technology Research Center Karamay834000 China

As artificial intelligence technology gains widespread adoption in biomedicine, the exploration of integrating biofluidic Raman spectroscopy for enhanced disease diagnosis opens up new prospects for the practical application of Raman spectroscopy in clinical settings. However, for Systemic Lupus Erythematosus (SLE), the relatively weak signals from a single Raman spectroscopic data make it difficult to obtain desirable classification results. Although the surface enhancement technique can enhance the scattered signal of Raman spectroscopic data, the sensitivity of the SERS substrate to airborne impurities and the non-uniform distribution of hotspots degrade some of the signals. To fully utilize both types of data, this paper proposes a two-branch residual-attention network (DBRAN) fusion technique that allows the original spectra to complement the degraded portion, improving the model's classification accuracy. The residual module is used to extract features, which preserves the original features while extracting the deep ones. At the same time, the study incorporates the attention module in both the upper and lower branches to handle the weight assignment of the two modal features more efficiently. The experimental results demonstrated that both the low-level fusion method and the intermediate-level fusion method significantly improved the diagnostic accuracy of SLE disease classification when compared with a single modality, and the intermediate-level fusion DBRAN achieved 100% classification accuracy, sensitivity, and specificity. Compared with the raw Raman spectrum single mode and SERS single mode, the improvement is 10% and 7%, respectively. This experiment realized the rapid diagnosis of SLE disease by fusing the spectral data of multiple modes, which provides a reference idea in the Raman spectroscopy field. © 2023, The Authors. All rights reserved.

关键词： Raman spectroscopy

Sle Diagnosis Research Based on Sers Combined with a Multi-Modal Fusion Method

学校读者我要写书评

暂无评论

SSRN 2023年

关键词： Raman spectroscopy

Dynamic Prefix-Tuning for Generative Template-based Event Extraction

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Liu, Xiao Huang, Heyan Shi, Ge Wang, Bo School of Computer Science and Technology Beijing Institute of Technology China Beijing Engineering Research Center of High Volume Language Information Processing and Cloud Computing Applications China Key Laboratory of Intelligent Information Processing and Information Security Ministry of Industry and Information Technology China Southeast Academy of Information Technology Beijing Institute of Technology China Faculty of Information Technology Beijing University of Technology China

We consider event extraction in a generative manner with template-based conditional generation. Although there is a rising trend of casting the task of event extraction as a sequence generation problem with prompts, these generation-based methods have two significant challenges, including using suboptimal prompts and static event type information. In this paper, we propose a generative template-based event extraction method with dynamic prefix (GTEE-DYNPREF) by integrating context information with type-specific prefixes to learn a context-specific prefix for each context. Experimental results show that our model achieves competitive results with the state-of-the-art classification-based model ONEIE on ACE 2005 and achieves the best performances on ERE. Additionally, our model is proven to be portable to new types of events effectively. Copyright © 2022, The Authors. All rights reserved.

关键词： Extraction

Gleffn: A Global-Local Event Feature Fusion Network for Micro-Expression Recognition

学校读者我要写书评

暂无评论

SSRN 2023年

作者： Guo, Cunhan Huang, Heyan School of Emergency Management Science and Engineering University of Chinese Academy of Science Beijing China Southeast Academy of information Technology Beijing Institute of Technology Putian China Beijing Engineering Research Center of High Volume Language Information Processing and Cloud Computing Applications China

Micro-expressions are brief and subtle facial movements that reveal a person’s true emotions. However, the low frame rate of conventional cameras limits the performance of micro-expression recognition (MER) systems. Event-based cameras are a novel technology that offer high frame rate and low latency, making them suitable for MER tasks, but they are hard to obtain. In this article, we propose a local event feature called local count image, which is derived from up-sampled data using the SloMo method. We also design a global-local event feature fusion network that combines local count image and global features extracted by dense optical flow to learn more discriminative features and solve MER tasks. Experiments showed that our method surpasses state-ofthe-art methods on several datasets. The ablation studies confirm the effectiveness of the local count image feature and the random data augmentation. To the best of our knowledge, our work is the first to tackle MER tasks from an event perspective, and it can facilitate the adoption of event cameras and provide some inspiration for future related works. © 2023, The Authors. All rights reserved.

关键词： Cameras

Capturing Global Structural information in Long Document Question Answering with Compressive Graph Selector Network

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Nie, Yuxiang Huang, Heyan Wei, Wei Mao, Xian-Ling School of Computer Science and Technology Beijing Institute of Technology China Beijing Engineering Research Center of High Volume Language Information Processing and Cloud Computing Applications China Beijing Institute of Technology Southeast Academy of Information Technology China Huazhong University of Science and Technology China

Long document question answering is a challenging task due to its demands for complex reasoning over long text. Previous works usually take long documents as non-structured flat texts or only consider the local structure in long documents. However, these methods usually ignore the global structure of the long document, which is essential for long-range understanding. To tackle this problem, we propose Compressive Graph Selector Network (CGSN) to capture the global structure in a compressive and iterative manner. Specifically, the proposed model consists of three modules: local graph network, global graph network and evidence memory network. Firstly, the local graph network builds the graph structure of the chunked segment in token, sentence, paragraph and segment levels to capture the short-term dependency of the text. Secondly, the global graph network selectively receives the information of each level from the local graph, compresses them into the global graph nodes and applies graph attention into the global graph nodes to build the long-range reasoning over the entire text in an iterative way. Thirdly, the evidence memory network is designed to alleviate the redundancy problem in the evidence selection via saving the selected result in the previous steps. Extensive experiments show that the proposed model outperforms previous methods on two datasets.1 Copyright © 2022, The Authors. All rights reserved.

关键词： Iterative methods

AttenWalker: Unsupervised Long-Document Question Answering via Attention-based Graph Walking

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Nie, Yuxiang Huang, Heyan Wei, Wei Mao, Xian-Ling School of Computer Science and Technology Beijing Institute of Technology China Beijing Engineering Research Center High Volume Language Information Processing and Cloud Computing Applications China Beijing Institute of Technology Southeast Academy of Information Technology China Huazhong University of Science and Technology China

Annotating long-document question answering (long-document QA) pairs is time-consuming and expensive. To alleviate the problem, it might be possible to generate long-document QA pairs via unsupervised question answering (UQA) methods. However, existing UQA tasks are based on short documents, and can hardly incorporate long-range information. To tackle the problem, we propose a new task, named unsupervised long-document question answering (ULQA), aiming to generate high-quality long-document QA instances in an unsupervised manner. Besides, we propose AttenWalker, a novel unsupervised method to aggregate and generate answers with long-range dependency so as to construct long-document QA pairs. Specifically, AttenWalker is composed of three modules, i.e., span collector, span linker and answer aggregator. Firstly, the span collector takes advantage of constituent parsing and reconstruction loss to select informative candidate spans for constructing answers. Secondly, by going through the attention graph of a pre-trained long-document model, potentially interrelated text spans (that might be far apart) could be linked together via an attention-walking algorithm. Thirdly, in the answer aggregator, linked spans are aggregated into the final answer via the mask-filling ability of a pre-trained model. Extensive experiments show that AttenWalker outperforms previous methods on Qasper and NarrativeQA. In addition, AttenWalker also shows strong performance in the few-shot learning setting. Copyright © 2023, The Authors. All rights reserved.

关键词：