检索结果-内蒙古大学图书馆

31st International Conference on Computational Linguistics, COLING 2025

作者： Guo, Xiaotao Li, Jiang Su, Xiangdong Zhang, Fujun College of Computer Science Inner Mongolia University Hohhot China National & Local Joint Engineering Research Center of Intelligent Information Processing Technology for Mongolian Hohhot China Inner Mongolia Key Laboratory of Multilingual Artificial Intelligence Technology Hohhot China

ISBN: (纸本)9798891761964

Sentence ordering is the task of rearranging a set of unordered sentences into a coherent and logically consistent sequence. Recent work has primarily used pre-trained language models, achieving significant success in the task. However, existing sentence ordering corpora are predominantly in English, and comprehensive benchmark datasets for non-English languages are unavailable. Meanwhile, current datasets often insert specific markers into paragraphs, inadvertently making the logical sequence between sentences more apparent and reducing the models' ability to handle genuinely unordered sentences in real applications. To address these limitations, we develop C3LRSO, a high-quality Chinese sentence ordering dataset that overcomes the aforementioned shortcomings by providing genuinely unordered sentences without artificial segmentation cues. Furthermore, given the outstanding performance of large language models on NLP tasks, we evaluate these models on our dataset for this task. Additionally, we propose a simple yet effective parameter-free approach that outperforms existing methods on this task. Experiments demonstrate the challenging nature of the dataset and the strong performance of our proposed method. These findings highlight the potential for further research in sentence ordering and the development of more robust language models. Our dataset is freely available at https://***/JasonGuo1/C3LRSO. © 2025 Association for Computational Linguistics.

关键词： Computational linguistics

来源：评论

学校读者我要写书评

暂无评论

Distance-Adaptive Quaternion Knowledge Graph Embedding with Bidirectional Rotation 31

Distance-Adaptive Quaternion Knowledge Graph Embedding with ...

引用

31st International Conference on Computational Linguistics, COLING 2025

作者： Wang, Weihua Liang, Qiuyu Bao, Feilong Gao, Guanglai College of Computer Science Inner Mongolia University Hohhot China National and Local Joint Engineering Research Center of Intelligent Information Processing Technology for Mongolian Hohhot China Inner Mongolia Key Laboratory of Multilingual Artificial Intelligence Technology Hohhot China

ISBN: (纸本)9798891761964

Quaternion contains one real part and three imaginary parts, which provided a more expressive hypercomplex space for learning knowledge graph. Existing quaternion embedding models measure the plausibility of a triplet through either semantic matching or geometric distance scoring functions. However, it appears that semantic matching diminishes the separability of entities, while the distance scoring function weakens the semantics of entities. To address this issue, we propose a novel quaternion knowledge graph embedding model. Our model combines semantic matching with the geometric distance of entities to better measure the plausibility of triplets. Specifically, in the quaternion space, we perform a right rotation on head entity and a reverse rotation on tail entity to learn rich semantic features. We then utilize distance-adaptive translations to learn geometric distance between entities. Furthermore, we provide mathematical proofs to demonstrate our model can handle complex logical relationships. Extensive experimental results and analyses show our model significantly outperforms previous models on well-known knowledge graph completion benchmark datasets. Our code is available at https://***/llqy123/DaBR. © 2025 Association for Computational Linguistics.

关键词： Knowledge graph

来源：评论

学校读者我要写书评

暂无评论

A Lightweight Chinese Summarization Generation Method Based on Knowledge Distillation 4

A Lightweight Chinese Summarization Generation Method Based ...

引用

4th International Conference on information Science, Parallel and Distributed Systems, ISPDS 2023

作者： Liu, Shiyue Xiao, Yannan Wang, Xueyong Huang, Wenming Deng, Zhenrong School of Computer and Information Security Guilin University of Electronic Technology Guangxi Key Laboratory of Image and Graphic Intelligent Processing Guilin China School of Computer and Information Security Guilin University of Electronic Technology Guilin China Guilin Xintong Technology Co. Guilin China

ISBN: (纸本)9798350337181

In recent years, large-scale pretrained natural language processing models such as BERT, GPT3 have achieved good results in processing tasks. However, in daily applications, these large-scale language models usually exist large model size or long running time, and the model portability and application are not very convenient. To settle this problem, we come up with a lightweight summarization generation method based on knowledge distillation (LW-BERT-KD, abbreviated as LBK model) to prove that complex neural network knowledge (taking BERT model as an example) can be extracted to lightweight language processing models (taking BiLSTM model as an example), thus achieving the purpose of complex knowledge extraction and model compression. In this paper, the BERT model is regarded as the teacher model, and the BiLSTM is regarded as the student model. The knowledge distillation technology is the transfer of knowledge from teacher models to student models, so that the generated model is effective, lightweight, and easy to port. On the Chinese abstract generation dataset, the training effect of this model is significantly better than the basic Transformer baseline. The experimental results show that if the training parameters are increased by 100 times, the model achieves comparable results with complex preprocessing language models. © 2023 IEEE.

关键词： Distillation

来源：评论

学校读者我要写书评

暂无评论

End-to-End Paired Ambisonic-Binaural Audio Rendering

引用

IEEE/CAA Journal of Automatica Sinica 2024年第2期11卷 502-513页

作者： Yin Zhu Qiuqiang Kong Junjie Shi Shilei Liu Xuzhou Ye Ju-Chiang Wang Hongming Shan Junping Zhang Shanghai Key Laboratory of Intelligent Information Processing School of Computer ScienceFudan UniversityShanghai 200433China Beijing ByteDance Technology Co.Ltd. Shanghai 201102China. Chinese University of Hong Kong Hong KongChina IEEE Institute of Science and Technology for Brain-Inspired Intelligence and MOE Frontiers Center for Brain Science Fudan UniversityShanghai 200433 Shanghai Center for Brain Science and Brain-Inspired Technology Shanghai 200031China

Binaural rendering is of great interest to virtual reality and immersive media. Although humans can naturally use their two ears to perceive the spatial information contained in sounds, it is a challenging task for machines to achieve binaural rendering since the description of a sound field often requires multiple channels and even the metadata of the sound sources. In addition, the perceived sound varies from person to person even in the same sound field. Previous methods generally rely on individual-dependent head-related transferred function(HRTF)datasets and optimization algorithms that act on HRTFs. In practical applications, there are two major drawbacks to existing methods. The first is a high personalization cost, as traditional methods achieve personalized needs by measuring HRTFs. The second is insufficient accuracy because the optimization goal of traditional methods is to retain another part of information that is more important in perception at the cost of discarding a part of the information. Therefore, it is desirable to develop novel techniques to achieve personalization and accuracy at a low cost. To this end, we focus on the binaural rendering of ambisonic and propose 1) channel-shared encoder and channel-compared attention integrated into neural networks and 2) a loss function quantifying interaural level differences to deal with spatial information. To verify the proposed method, we collect and release the first paired ambisonic-binaural dataset and introduce three metrics to evaluate the content information and spatial information accuracy of the end-to-end methods. Extensive experimental results on the collected dataset demonstrate the superior performance of the proposed method and the shortcomings of previous methods.

关键词： Ambisonic attention binaural rendering neural network

来源：评论

学校读者我要写书评

暂无评论

Continual Learning with a Memory of Non-similar Samples 1

引用

17th International Conference on Bio-Inspired Computing: Theories and Applications, BIC-TA 2022

作者： Min, Qilang He, Juanjuan Yang, Liuyan Fu, Yue College of Computer Science and Technology Wuhan University of Science and Technology Wuhan430081 China Hubei Province Key Laboratory of Intelligent Information Processing and Real-Time Industrial System Wuhan China

ISBN: (数字)9789819915491

ISBN: (纸本)9789819915484

The replay method is an effective strategy to address the problem of catastrophic forgetting, a key challenge for continuous learning. However, the sample sets obtained by replay-based methods generally suffer from local data information deficiencies. This problem leads to an imbalance in the plasticity-stability of the model on older tasks. This paper proposes a novel method, called non-similar sample storage (NSS). Non-similar refers to the Euclidean distance for feature vectors of different samples being far. NSS extracts the feature vectors of the samples and then calculates the similarity of the feature vectors for each sample after the current task training. Samples that contribute less to the model classification effect among similar samples are iteratively deleted, and the subset of low-similarity samples is retained. Moreover, NSS reserves 30% of the storage space for saving samples near the center of the sample set. Low-similarity samples stored by NSS get larger losses during the replay process, leading to lower training effectiveness of the current task. This paper introduces a knowledge distillation strategy to solve this problem. A variable parameter was used to balance the classification loss of the new task with the distillation loss of the old task (NSS-D). Experimental results in CIFAR10 and imbalanced CIFAR10 show that NSS maximizes the data’s global information and can retain the model’s ability to recognize old tasks better. Compared with classical algorithms, NSS-D performs better on CIFAR100 (48.8%) and ImageNet-200 (36.7%). © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Distillation

来源：评论

学校读者我要写书评

暂无评论

Consistency-aware unsupervised label learning for cross-domain person re-identification

引用

Multimedia Tools and Applications 2024年 1-20页

作者： Geng, Yanbing Lian, Yongjian Cui, Fangshu Zhang, Xiaowei Zhou, Mingliang Zhang, Geao School of Data Science and Technology North University of China Shanxi Taiyuan030051 China Shandong Key Laboratory of Intelligent Information Processing School of Computer Science and Technology Qingdao University Qingdao266071 China School of Computer Science Chongqing University 174 Shazheng Street Shapingba District Chongqing400044 China College of Information Science and Engineering Northeastern University 11 Lane 3 Wenhua Road Heping District Liaoning Shenyang110819 China

Un-refined pseudo labels always disturb the cross-domain Re-ID performance in unsupervised clustering methods. In this paper, we propose a consistency-aware unsupervised label learning network to refine noisy labels for promising domain adaptation. Specifically, we use a hetero-generated label consistency constraint to mine possible label information, keep a reference consistency of the target pairs on the relative comparative characteristics. We then use a cross-granularity attention consistency constraint to refine the pseudo labels meanwhile learning multi-granularity discriminative representation, keeping a self-similarity consistency of the corresponding parts on the discriminative regions. Our model is fine-tuned by the refined pseudo-labels to reduce the domain gap while coping with the intra-domain variation. Experimental results demonstrate that the proposed method can achieve superior performance on several benchmark datasets. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024.

关键词： Benchmarking

来源：评论

学校读者我要写书评

暂无评论

Denoising Diffusion Path: Attribution Noise Reduction with An Auxiliary Diffusion Model 38

Denoising Diffusion Path: Attribution Noise Reduction with A...

引用

38th Conference on Neural information processing Systems, NeurIPS 2024

作者： Lei, Yiming Li, Zilong Zhang, Junping Shan, Hongming Shanghai Key Laboratory of Intelligent Information Processing School of Computer Science Fudan University China Institute of Science and Technology for Brain-Inspired Intelligence MOE Key Laboratory of Computational Neuroscience and Brain-Inspired Intelligence MOE Frontiers Center for Brain Science Fudan University China

The explainability of deep neural networks (DNNs) is critical for trust and reliability in AI systems. Path-based attribution methods, such as integrated gradients (IG), aim to explain predictions by accumulating gradients along a path from a baseline to the target image. However, noise accumulated during this process can significantly distort the explanation. While existing methods primarily concentrate on finding alternative paths to circumvent noise, they overlook a critical issue: intermediate-step images frequently diverge from the distribution of training data, further intensifying the impact of noise. This work presents a novel Denoising Diffusion Path (DDPath) to tackle this challenge by harnessing the power of diffusion models for denoising. By exploiting the inherent ability of diffusion models to progressively remove noise from an image, DDPath constructs a piece-wise linear path. Each segment of this path ensures that samples drawn from a Gaussian distribution are centered around the target image. This approach facilitates a gradual reduction of noise along the path. We further demonstrate that DDPath adheres to essential axiomatic properties for attribution methods and can be seamlessly integrated with existing methods such as IG. Extensive experimental results demonstrate that DDPath can significantly reduce noise in the attributions-resulting in clearer explanations-and achieves better quantitative results than traditional path-based methods. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

OSAP‐Loss:Efficient optimization of average precision via involving samples after positive ones towards remote sensing image retrieval

引用

CAAI Transactions on Intelligence technology 2023年第4期8卷 1191-1212页

作者： Xin Yuan Xin Xu Xiao Wang Kai Zhang Liang Liao Zheng Wang Chia‐Wen Lin School of Computer Science and Technology Wuhan University of Science and TechnologyWuhanChina Hubei Province Key Laboratory of Intelligent Information Processing and Real‐time Industrial System Wuhan University of Science and TechnologyWuhanChina Nanyang Technological University SingaporeSingapore School of Computer Science Wuhan UniversityWuhanChina Department of Electrical Engineering Industrial Technology Research InstituteNational Tsing Hua UniversityHsinchuChina

In existing remote sensing image retrieval(RSIR)datasets,the number of images among different classes varies dramatically,which leads to a severe class imbalance *** studies propose to train the model with the ranking‐based metric(e.g.,average precision[AP]),because AP is robust to class ***,current AP‐based methods overlook an important issue:only optimising samples ranking before each positive sample,which is limited by the definition of AP and is prone to local *** achieve global optimisation of AP,a novel method,namely Optimising Samples after positive ones&AP loss(OSAP‐Loss)is proposed in this ***,a novel superior ranking function is designed to make the AP loss differentiable while providing a tighter upper ***,a novel loss called Optimising Samples after Positive ones(OSP)loss is proposed to involve all positive and negative samples ranking after each positive one and to provide a more flexible optimisation strategy for each ***,a graphics processing unit memory‐free mechanism is developed to thoroughly address the non‐decomposability of AP *** experimental results on RSIR as well as conventional image retrieval datasets show the superiority and competitive performance of OSAP‐Loss compared to the state‐of‐the‐art.

关键词： computer vision image retrieval metric learning

来源：评论

学校读者我要写书评

暂无评论

A Gradient-Guided Evolutionary Approach to Training Deep Neural Networks

引用

IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS 2022年第9期33卷 4861-4875页

作者： Yang, Shangshang Tian, Ye He, Cheng Zhang, Xingyi Tan, Kay Chen Jin, Yaochu Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education School of Computer Science and Technology Anhui University Hefei China Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education Institutes of Physical Science and Information Technology Anhui University Hefei China Department of Computer Science and Engineering Guangdong Provincial Key Laboratory of Brain-Inspired Intelligent Computation Southern University of Science and Technology Shenzhen China Department of Computing The Hong Kong Polytechnic University Hong Kong SAR Department of Computer Science University of Surrey Guildford U.K

It has been widely recognized that the efficient training of neural networks (NNs) is crucial to classification performance. While a series of gradient-based approaches have been extensively developed, they are criticized for the ease of trapping into local optima and sensitivity to hyperparameters. Due to the high robustness and wide applicability, evolutionary algorithms (EAs) have been regarded as a promising alternative for training NNs in recent years. However, EAs suffer from the curse of dimensionality and are inefficient in training deep NNs (DNNs). By inheriting the advantages of both the gradient-based approaches and EAs, this article proposes a gradient-guided evolutionary approach to train DNNs. The proposed approach suggests a novel genetic operator to optimize the weights in the search space, where the search direction is determined by the gradient of weights. Moreover, the network sparsity is considered in the proposed approach, which highly reduces the network complexity and alleviates overfitting. Experimental results on single-layer NNs, deep-layer NNs, recurrent NNs, and convolutional NNs (CNNs) demonstrate the effectiveness of the proposed approach. In short, this work not only introduces a novel approach for training DNNs but also enhances the performance of EAs in solving large-scale optimization problems.

关键词： Artificial neural networks Training Optimization Search problems Scalability Complexity theory Genetic algorithms Deep neural networks (DNNs) evolutionary algorithm (EA) genetic operator gradient descent sparsity

来源：评论

学校读者我要写书评

暂无评论

AUCSeg: AUC-oriented Pixel-level Long-tail Semantic Segmentation 38

AUCSeg: AUC-oriented Pixel-level Long-tail Semantic Segmenta...

引用

38th Conference on Neural information processing Systems, NeurIPS 2024

作者： Han, Boyu Xu, Qianqian Yang, Zhiyong Bao, Shilong Wen, Peisong Jiang, Yangbangyan Huang, Qingming Key Lab. of Intelligent Information Processing Institute of Computing Technology CAS China School of Computer Science and Tech. University of Chinese Academy of Sciences China Peng Cheng Laboratory China Key Laboratory of Big Data Mining and Knowledge Management CAS China

The Area Under the ROC Curve (AUC) is a well-known metric for evaluating instance-level long-tail learning problems. In the past two decades, many AUC optimization methods have been proposed to improve model performance under long-tail distributions. In this paper, we explore AUC optimization methods in the context of pixel-level long-tail semantic segmentation, a much more complicated scenario. This task introduces two major challenges for AUC optimization techniques. On one hand, AUC optimization in a pixel-level task involves complex coupling across loss terms, with structured inner-image and pairwise inter-image dependencies, complicating theoretical analysis. On the other hand, we find that mini-batch estimation of AUC loss in this case requires a larger batch size, resulting in an unaffordable space complexity. To address these issues, we develop a pixel-level AUC loss function and conduct a dependency-graph-based theoretical analysis of the algorithm's generalization ability. Additionally, we design a Tail-Classes Memory Bank (T-Memory Bank) to manage the significant memory demand. Finally, comprehensive experiments across various benchmarks confirm the effectiveness of our proposed AUCSeg method. The code is available at https://***/boyuh/AUCSeg. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

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：