检索结果-内蒙古大学图书馆

SSRN 2022年

作者： Sun, Jiayin Dong, Qiulei National Laboratory of Pattern Recognition CASIA Beijing100190 China School of Artificial Intelligence UCAS Beijing100049 China Center for Excellence in Brain Science and Intelligence Technology CAS Beijing100190 China

Open-set recognition (OSR) aims to simultaneously detect unknown-class samples and classify known-class samples. Most of the existing OSR methods are inductive methods, which generally suffer from the domain shift problem that the learned model from the known-class domain might be unsuitable for the unknown-class domain. Addressing this problem, inspired by the success of transductive learning for alleviating the domain shift problem in many other visual tasks, we propose an Iterative Transductive OSR framework, called IT-OSR, which implements three explored modules iteratively, including a reliability sampling module, a feature generation module, and a baseline update module. Specifically, at each iteration, a dual-space consistent sampling approach is presented in the explored reliability sampling module for selecting some relatively more reliable ones from the test samples according to their pseudo labels assigned by a baseline method, which could be an arbitrary inductive OSR method. Then, a conditional dual-adversarial generative network under an orthogonal coding condition is designed in the feature generation module to generate discriminative sample features of both known and unknown classes according to the selected test samples with their pseudo labels. Finally, the baseline method is updated for sample re-prediction in the baseline update module by jointly utilizing the generated features, the selected test samples with pseudo labels, and the training samples. Extensive experimental results on both the standard-dataset and the cross-dataset settings demonstrate that the derived transductive methods, by introducing two typical inductive OSR methods into the proposed IT-OSR framework, achieve better performances than 19 state-of-the-art methods in most cases. © 2022, The Authors. All rights reserved.

关键词： Iterative methods

来源：评论

学校读者我要写书评

暂无评论

Learning Latent Causal Factors from the Intricate Sensor Feedback of Contact-Rich Robotic Assembly Tasks

SSRN

引用

SSRN 2023年

作者： Chen, Yurou Yu, Jiyang Lin, Zhenyang Shen, Liancheng Liu, Zhiyong The State Key Lab of Management and Control for Complex Systems Institute of Automation Chinese Academy of Sciences Beijing China School of Artificial Intelligence University of Chinese Academy of Sciences Beijing China The Center for Excellence in Brain Science and Intelligence Technology Chinese Academy of Sciences Shanghai China Beijing Institute of Spacecraft System Engineering China Academy of Space Technology Beijing China

Achieving precise robotic assembly is paramount for safety and performance in industrial settings. Conventional assembly methods require tedious manual adjustment of many parameters, and it is challenging to meet the assembly requirements with tight clearance. Robot learning has been at the forefront of research, showing potential for automation and intelligence in robotic operations. However, intricate information such as force feedback is indispensable for high-precision robot assembly tasks, and current learning methods grapple with processing this complex observation data due to discontinuous force changes during operation and the inherent noise from the force sensor. To alleviate the low learning efficiency from the problematic observations, we propose introducing the concept of latent causal factors that cause the sensor observations and truly matter in assembly tasks. This paper analyses the impact of discontinuous and noisy observations and offers two ways to infer causal factors. Based on the implied latent factors, we propose a method that learns high-precision assembly policies from interacting with the environment. The algorithm's performance is evaluated in simulated and real-world nut assembly environments, demonstrating significant improvements over the previous methods. This research also underscores the promise of causal inference in addressing intricate industrial challenges%. © 2023, The Authors. All rights reserved.

关键词： Robotic assembly

来源：评论

学校读者我要写书评

暂无评论

Decoupling Pronunciation and Language for End-to-End Code-Switching Automatic Speech Recognition

Decoupling Pronunciation and Language for End-to-End Code-Sw...

引用

IEEE International Conference on Acoustics, Speech and Signal Processing

作者： Shuai Zhang Jiangyan Yi Zhengkun Tian Ye Bai Jianhua Tao Zhengqi wen School of Artificial Intelligence University of Chinese Academy of Sciences China NLPR Institute of Automation Chinese Academy of Sciences China CAS Center for Excellence in Brain Science and Intelligence Technology China

Despite the recent significant advances witnessed in end-to-end (E2E) ASR system for code-switching, hunger for audio-text paired data limits the further improvement of the models' performance. In this paper, we propose a decoupled transformer model to use mono-lingual paired data and unpaired text data to alleviate the problem of code-switching data shortage. The model is decoupled into two parts: audio-to-phoneme (A2P) network and phoneme-to-text (P2T) network. The A2P network can learn acoustic pattern scenarios using large-scale monolingual paired data. Meanwhile, it generates multiple phoneme sequence candidates for single audio data in real time during the training process. Then the generated phoneme-text paired data is used to train the P2T network. This network can be pre-trained with large amounts of external unpaired text data. By using monolingual data and unpaired text data, the decoupled transformer model reduces the high dependency on code-switching paired training data of E2E model to a certain extent. Finally, the two networks are optimized jointly through attention fusion. We evaluate the proposed method on the public Mandarin-English code-switching dataset. Compared with our transformer baseline, the proposed method achieves 18.14% relative mix error rate reduction.

关键词： Training Speech coding Error analysis Conferences Training data Signal processing Data models

来源：评论

学校读者我要写书评

暂无评论

EventMix: An Efficient Augmentation Strategy for Event-Based Data

arXiv

引用

arXiv 2022年

作者： Shen, Guobin Zhao, Dongcheng Zeng, Yi Research Center for Brain-Inspired Intelligence CASIA China Center for Excellence in Brain Science and Intelligence Technology CAS China National Laboratory of Pattern Recognition CASIA China School of Future Technology University of Chinese Academy of Sciences China School of Artificial Intelligence University of Chinese Academy of Sciences China

High-quality and challenging event stream datasets play an important role in the design of an efficient event-driven mechanism that mimics the brain. Although event cameras can provide high dynamic range and low-energy event stream data, the scale is smaller and more difficult to obtain than traditional frame-based data, which restricts the development of neuromorphic computing. Data augmentation can improve the quantity and quality of the original data by processing more representations from the original data. This paper proposes an efficient data augmentation strategy for event stream data: EventMix. We carefully design the mixing of different event streams by Gaussian Mixture Model to generate random 3D masks and achieve arbitrary shape mixing of event streams in the spatio-temporal dimension. By computing the relative distances of event streams, we propose a more reasonable way to assign labels to the mixed samples. The experimental results on multiple neuromorphic datasets have shown that our strategy can improve its performance on neuromorphic datasets both for ANNs and SNNs, and we have achieved state-of-the-art performance on DVS-CIFAR10, N-Caltech101, N-CARS, and DVS-Gesture datasets. Copyright © 2022, The Authors. All rights reserved.

关键词： Mixing

来源：评论

学校读者我要写书评

暂无评论

Non-invasive magnetocardiography of living rat based on diamond quantum sensor

arXiv

引用

arXiv 2024年

作者： Yu, Ziyun Xie, Yijin Jin, Guodong Zhu, Yunbin Zhang, Qi Shi, Fazhan Wan, Fang-Yan Luo, Hongmei Tang, Ai-Hui Rong, Xing CAS Key Laboratory of Microscale Magnetic Resonance School of Physical Sciences University of Science and Technology of China Hefei230026 China CAS Center for Excellence in Quantum Information and Quantum Physics University of Science and Technology of China Hefei230026 China Institute of Quantum Sensing School of Physics Zhejiang University Hangzhou310027 China Hefei National Laboratory University of Science and Technology of China Hefei230088 China CAS Key Laboratory of Brain Function and Disease Hefei National Laboratory for Physical Sciences at the Microscale Division of Life Sciences and Medicine University of Science and Technology of China Hefei230026 China Institute of Artificial Intelligence Hefei Comprehensive National Science Center Hefei230088 China

Magnetocardiography (MCG) has emerged as a sensitive and precise method to diagnose cardiovascular diseases, providing more diagnostic information than traditional technology. However, the sensor limitations of conventional MCG systems, such as large size and cryogenic requirement, have hindered the widespread application and in-depth understanding of this technology. In this study, we present a high-sensitivity, room-temperature MCG system based on the negatively charged Nitrogen-Vacancy (NV) centers in diamond. The magnetic cardiac signal of a living rat, characterized by an approximately 20 pT amplitude in the R-wave, is successfully captured through non-invasive measurement using this innovative solid-state spin sensor. To detect these extremely weak biomagnetic signals, we utilize sensitivity-enhancing techniques such as magnetic flux concentration. These approaches have enabled us to simultaneously achieve a magnetometry sensitivity of 9 pT · Hz-1/2 and a sensor scale of 5 mm. By extending the sensing scale of the NV centers from cellular and molecular level to macroscopic level of living creatures, we have opened the future of solid-state quantum sensing technologies in clinical environments. © 2024, CC BY.

关键词： Magnetocardiography

来源：评论

学校读者我要写书评

暂无评论

Hydrogen peroxide in midbrain sleep neurons regulates sleep homeostasis

引用

Cell Metabolism 2025年第6期37卷 1442-1452.e7页

作者： Tian, Yujing Kang, Luwei Ha, Ngoc T. Deng, Juan Liu, Danqian Center for Excellence in Brain Science and Intelligence Technology Chinese Academy of Sciences Shanghai 200031 China University of Chinese Academy of Sciences Beijing 100049 China Department of Anesthesiology Huashan Hospital State Key Laboratory of Medical Neurobiology Institute for Translational Brain Research MOE Frontiers Center for Brain Science Fudan University Shanghai 200032 China

Sleep could protect animals from oxidative damage, yet the dynamic interplay between the redox state and sleep homeostasis remains unclear. Here, we show that acute sleep deprivation (SD) in mice caused a general increase in brain oxidation, particularly in sleep-promoting regions. In vivo imaging of intracellular hydrogen peroxide (H₂O₂) real-time dynamics revealed that in nigra sleep neurons, the increase in cytosolic but not mitochondrial H₂O₂ reflects sleep debt and tracks spontaneous wakefulness by positively correlating with wake duration. By controllably manipulating intraneuronal H₂O₂, we discovered that H₂O₂ elevation is required for compensatory sleep and causally promotes sleep initiation, at least partly dependent on transient receptor potential melastatin 2 (TRPM2) channel. However, excessive H₂O₂ induced brain inflammation and sleep fragmentation. Together, our study demonstrates intraneuronal H₂O₂ as a crucial signaling molecule that translates brain redox imbalance into sleep drive and underscores the significance of oxidative eustress in sleep homeostasis. © 2025 Elsevier Inc.

关键词： chemogenetic H2O2 H2O2 generation hydrogen peroxide in vivo H2O2 imaging oxidative eustress redox signaling sleep circuits sleep homeostasis TRPM2

来源：评论

学校读者我要写书评

暂无评论

MultimodalSleepNet: A Lightweight Neural Network Model for Sleep Staging based on Multimodal Physiological Signals

MultimodalSleepNet: A Lightweight Neural Network Model for S...

引用

IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

作者： Kunbo Cui Mingqi Zhao Minxin He Di Liu Qinglin Zhao Bin Hu School of Information Science and Engineering Lanzhou University Lanzhou China Key Laboratory of Special Functional Materials and Structural Design Ministry of Education Lanzhou University Lanzhou China School of Medical Technology Beijing Institute of Technology Beijing China CAS Center for Excellence in Brain Science and Intelligence Technology Shanghai Institutes for Biological Sciences Chinese Academy of Sciences Shanghai China Joint Research Center for Cognitive Neurosensor Technology of Lanzhou University & Institute of Semiconductors Chinese Academy of Sciences Lanzhou China

ISBN: (数字)9798350386226

ISBN: (纸本)9798350386233

With social development, the demand for automatic sleep quality assessment in wearable devices is increasing, especially as sleep quality is closely related to the diagnosis of psychiatric disorders. However, existing automatic sleep stage classification models are mostly designed for unimodal signals, with an emphasis on increasing parameter scale and model depth. As a result, it is challenging to meet the requirements for both lightweight models and high accuracy in wearable device-based automatic sleep staging tasks. To address this problem, this study introduces a novel lightweight model for sleep staging, MultimodalSleepNet, which is based on multi-modal physiological signals. Specifically, the model is designed to capture the temporal dynamics of physiological signals and the spatial interactions between multimodal signals. Additionally, an inflationary convolution mechanism is incorporated to accelerate temporal feature extraction. We validate the model using the publicly available Sleep-EDF-Expanded dataset. Compared to similar studies, our model achieves outstanding performance, with accuracies of 93.1% and 90.2% in the three-stage and five-stage sleep recognition tasks, respectively. Notably, the three-stage classification results show an 11.9% improvement in modal fusion accuracy compared to unimodal signals, with an 8.9% improvement in multiclass F1 score and a 20.8% increase in Cohen’s kappa coefficient. In conclusion, our study offers a reference for the design of lightweight models for sleep staging and provides new insights into feature extraction and fusion of multimodal signals.

关键词： Manifolds Accuracy Convolution Neural networks Feature extraction Physiology Quality assessment Data mining Wearable devices Covariance matrices

来源：评论

学校读者我要写书评

暂无评论

Heterogeneous Effects of Eye-close and Eye-open on Electroencephalographic Microstates of Depressed brain in Resting State

Heterogeneous Effects of Eye-close and Eye-open on Electroen...

引用

IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

作者： Kunbo Cui Mingqi Zhao Di Liu Minxin He Qinglin Zhao Bin Hu School of Information Science and Engineering Lanzhou University Lanzhou China Key Laboratory of Special Functional Materials and Structural Design Ministry of Education Lanzhou University Lanzhou China School of Medical Technology Beijing Institute of Technology Beijing China CAS Center for Excellence in Brain Science and Intelligence Technology Shanghai Institutes for Biological Sciences Chinese Academy of Sciences Shanghai China Joint Research Center for Cognitive Neurosensor Technology of Lanzhou University & Institute of Semiconductors Chinese Academy of Sciences Lanzhou China

ISBN: (数字)9798350386226

ISBN: (纸本)9798350386233

Previous studies have shown that the resting-state electroencephalogram (EEG) of depressed patients exhibits abnormal dynamic activation of large-scale networks in microstate analysis. However, the problem of heterogeneous results in microstate features as physiological markers for depressive disorders has not been addressed. An important factor contributing to this problem is that previous studies have overlooked the effects of eye-opening and eye-closing on resting-state EEG microstates in depression. To address this gap, the present study proposes a new microstate delineation method to accurately identify EEG microstates in both open-eye and closed-eye resting states, and further analyzes the differential performance of the depressed group and the control group in the two states. We validated our method on a dataset containing 64-channel EEG data (55 cases in the depressed group and 55 cases in the control group). The results showed that the classical seven microstate topographies were present in both open-eye and closed-eye conditions, which may explain why other studies have overlooked the impact of eye state on resting-state microstates in depression. In addition, the microstate characteristics of depression were differentially expressed in the open-eye and closed-eye conditions, with the depressed group showing more significant results in the closed-eye condition. Overall, our study demonstrates that eye state affects depression microstates and provides new insights to address the problem of heterogeneity in the results of EEG microstate studies of depression.

关键词： Frontal lobe Surfaces Depression Electroencephalography Physiology Bioinformatics

来源：评论

学校读者我要写书评

暂无评论

Changes of Effective Connectivity in the Alpha Band Characterize Differential Processing of Audiovisual Information in Cross-Modal Selective Attention

引用

Neuroscience Bulletin 2020年第9期36卷 1009-1022页

作者： Weikun Niu Yuying Jiang Xin Zhang Tianzi Jiang Yujin Zhang Shan Yu Brainnetome Center Institute of AutomationChinese Academy of SciencesBeijing 100190China National Laboratory of Pattern Recognition Institute of AutomationChinese Academy of SciencesBeijing 100190China Center for Excellence in Brain Science and Intelligence Technology Chinese Academy of SciencesBeijing 100190China University of Chinese Academy of Sciences Beijing 100049China

Cross-modal selective attention enhances the processing of sensory inputs that are most relevant to the task at *** differential processing could be mediated by a swift network reconfiguration on the macroscopic level,but this remains a poorly understood *** tackle this issue,we used a behavioral paradigm to introduce a shift of selective attention between the visual and auditory domains,and recorded scalp electroencephalographic signals from eight healthy *** changes in effective connectivity caused by the cross-modal attentional shift were delineated by analyzing spectral Granger Causality(GC),a metric of frequency-specific effective *** data-driven methods of pattern-classification and feature-analysis,we found that a change in the a band(12 Hz-15 Hz) of GC is a stable feature across different individuals that can be used to decode the attentional ***,auditory attention induces more pronounced information flow in the α band,especially from the parietal-occipital areas to the temporal-parietal areas,compared to the case of visual attention,reflecting a reconfiguration of interaction in the macroscopic brain network accompanying different *** results support the role of α oscillation in organizing the information flow across spatially-separated brain areas and,thereby,mediating cross-modal selective attention.

关键词： Human EEG Audiovisual selective attention Granger Causality Pattern classification

来源：评论

学校读者我要写书评

暂无评论

Multi-scale Evolutionary Neural Architecture Search for Deep Spiking Neural Networks

arXiv

引用

arXiv 2023年

作者： Pan, Wenxuan Zhao, Feifei Shen, Guobin Han, Bing Zeng, Yi The Brain-inspired Cognitive Intelligence Lab Institute of Automation Chinese Academy of Sciences Beijing100190 China School of Artificial Intelligence University of Chinese Academy of Sciences Beijing100049 China School of Future Technology University of Chinese Academy of Sciences Beijing100049 China University of Chinese Academy of Sciences Beijing100049 China Center for Excellence in Brain Science and Intelligence Technology Chinese Academy of Sciences Shanghai200031 China

Spiking Neural Networks (SNNs) have received considerable attention not only for their superiority in energy efficiency with discrete signal processing but also for their natural suitability to integrate multi-scale biological plasticity. However, most SNNs directly adopt the structure of the well-established Deep Neural Networks (DNNs), and rarely automatically design Neural Architecture Search (NAS) for SNNs. The neural motifs topology, modular regional structure and global cross-brain region connection of the human brain are the product of natural evolution and can serve as a perfect reference for designing brain-inspired SNN architecture. In this paper, we propose a Multi-Scale Evolutionary Neural Architecture Search (MSE-NAS) for SNN, simultaneously considering micro-, meso- and macro-scale brain topologies as the evolutionary search space. MSE-NAS evolves individual neuron operation, self-organized integration of multiple circuit motifs, and global connectivity across motifs through a brain-inspired Indirect Evaluation (BIE) Function. This training-free fitness function could greatly reduce computational consumption and NAS's time, and its task-independent property enables the searched SNNs to exhibit excellent transferability on multiple datasets. Furthermore, MSE-NAS show robustness against the training method and noise. Extensive experiments demonstrate that the proposed algorithm achieves state-of-the-art (SOTA) performance with shorter simulation steps on static datasets (CIFAR10, CIFAR100) and neuromorphic datasets (CIFAR10-DVS and DVS128-Gesture). The thorough analysis also illustrates the significant performance improvement and consistent bio-interpretability deriving from the topological evolution at different scales and the BIE function. © 2023, CC BY-NC-ND.

关键词： Topology

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：