检索结果-内蒙古大学图书馆

31st International Conference on Computational Linguistics, COLING 2025

作者： Yang, Li Xiao, Zhiding Huang, Wenxin Zhong, Xian Hubei Key Laboratory of Big Data Intelligent Analysis and Application School of Computer Science and Information Engineering Hubei University China Hubei Key Laboratory of Transportation Internet of Things School of Computer Science and Artificial Intelligence Wuhan University of Technology China

ISBN: (纸本)9798891761964

The rapid development of multi-modal large language models (MLLMs) has positioned visual storytelling as a crucial area in content creation. However, existing models often struggle to maintain temporal, spatial, and narrative coherence across image sequences and frequently lack the depth and engagement of human-authored stories. To address these challenges, we propose Story with Large Language and Vision Assistant (StoryLLaVA), a novel framework for enhancing visual storytelling. Our approach introduces a Topic-Driven Narrative Optimizer (TDNO) that improves both the training data and MLLM models by integrating image descriptions, topic generation, and GPT-4-based refinements. Furthermore, we employ a preference-based ranked story sampling method that aligns model outputs with human storytelling preferences through positive-negative pairing. These two phases of the framework differ in their training methods: the former uses supervised fine-tuning, while the latter incorporates reinforcement learning with positive and negative sample pairs. Experimental results demonstrate that StoryLLaVA outperforms current models in visual relevance, coherence, and fluency, with LLM-based evaluations confirming the generation of richer and more engaging narratives. The enhanced dataset and model are available at https://***/XxxZzD/StoryLLaVA. © 2025 Association for Computational Linguistics.

关键词：

来源：评论

学校读者我要写书评

暂无评论

LLM Knows Geometry Better than Algebra: Numerical Understanding of LLM-Based Agents in A Trading Arena

arXiv

引用

arXiv 2025年

作者： Ma, Tianmi Du, Jiawei Huang, Wenxin Wang, Wenjie Xie, Liang Zhong, Xian Zhou, Joey Tianyi Hubei Key Laboratory of Transportation Internet of Things Wuhan University of Technology China Hubei Key Laboratory of Big Data Intelligent Analysis and Application Hubei University China Centre for Frontier AI Research Agency for Science Technology and Research Singapore Institute of High Performance Computing Agency for Science Technology and Research Singapore School of Computing National University of Singapore Singapore School of Science Wuhan University of Technology China

Recent advancements in large language models (LLMs) have significantly improved performance in natural language processing tasks. However, their ability to generalize to dynamic, unseen tasks, particularly in numerical reasoning, remains a challenge. Existing benchmarks mainly evaluate LLMs on problems with predefined optimal solutions, which may not align with real-world scenarios where clear answers are absent. To bridge this gap, we design the Agent Trading Arena, a virtual numerical game simulating complex economic systems through zero-sum games, where agents invest in stock portfolios. Our experiments reveal that LLMs, including GPT-4o, struggle with algebraic reasoning when dealing with plain-text stock data, often focusing on local details rather than global trends. In contrast, LLMs perform significantly better with geometric reasoning when presented with visual data, such as scatter plots or K-line charts, suggesting that visual representations enhance numerical reasoning. This capability is further improved by incorporating the reflection module, which aids in the analysis and interpretation of complex data. We validate our findings on NASDAQ STOCK dataset, where LLMs demonstrate stronger reasoning with visual data compared to text. Our code and data are publicly available at https://***/wekjsdvnm/***. © 2025, CC BY.

关键词： Natural language processing systems

来源：评论

学校读者我要写书评

暂无评论

Salix-Leaf: Find Main Veins of Signal Clusters for Practical Parallel Decoding

引用

IEEE Transactions on Mobile Computing 2025年

作者： Li, Yajun Zhao, Jumin Li, Dengao Wu, Hejun Xu, Shuang Bai, Ruiqin Taiyuan University of Technology College of Electronic Information and Optical Engineering Taiyuan China Key Laboratory of Big Data Fusion Analysis and Application of Shanxi Province Taiyuan China Intelligent Perception Engineering Technology Center of Shanxi Taiyuan China Taiyuan University of Technology College of Computer Science and Technology College of Data Science Taiyuan China Sun Yat-sen University Department of Computer Science Guangzhou China

Parallel decoding of backscatter improves communication throughput by enabling concurrent transmission of backscatter tags. In practical applications of parallel decoding, it is extremely difficult to distinguish collided signals in superclusters where multiple signal clusters overlap. Existing methods are usually effective for superclusters with uniformly distributed signals. Nevertheless, there are many more scenarios in which signals in superclusters tend to gather unevenly, and existing methods cannot work. Such uneven clustering of signals occurs due to the following two possible causes: (1) signal-strengthdifferences (SSDs) among tags;or (2) cluster drifting (CD) driven by interferences from other objects within communication environments. This paper proposes a novel scheme called SalixLeaf, which aims to identify the main veins of signal clusters to address this problem of superclusters with unevenly distributed signals. Salix-Leaf identifies the main vein of each signal cluster for fine-grained clustering so that the direction of the main veins can be used to verify the accuracy of clustering. In addition, SalixLeaf employs a supercluster decomposer that divides signals into different segments for clustering analysis, enhancing robustness and practicability. Experimental results show that Salix-Leaf achieves a 1.2-fold increase in throughput and a 25% reduction in bit error rate (BER) compared to the state-of-the-art. © 2002-2012 IEEE.

关键词： Backscattering

来源：评论

学校读者我要写书评

暂无评论

Sequence-to-Sequence Multi-Hop Knowledge Reasoning Based on Retentive Network 7

Sequence-to-Sequence Multi-Hop Knowledge Reasoning Based on ...

引用

7th International Conference on Computer Information Science and application Technology, CISAT 2024

作者： Su, Ruilin Zhang, Wanshan Yu, Dunhui Hubei University Wuhan China Wuhan430062 China Hubei University School of Computer Science and Information Engineering Wuhan430062 China Hubei University Hubei Key Laboratory of Big Data Intelligent Analysis and Application Wuhan430062 China Engineering and Technical Research Center of Hubei Province in Software Engineering Wuhan430062 China

ISBN: (纸本)9798350375107

Multi-hop Knowledge Reasoning is a task that involves generating an answer given a query and a knowledge graph. Existing sequence-to-sequence reasoning models use the Transformer to encode and decode sequences, but these models have some flaws, such as the inability to effectively handle long-sequence reasoning and susceptibility to exposure bias. To address these issues, we provides a sequence-to-sequence reasoning model named STSR, which is based on Retentive Network. Aiming at improving training efficiency through parallel training, this model leverages the advantages in Retentive Network, reduce time overhead, and enhance reasoning efficiency through iterative reasoning. It effectively mitigates the problems of high spatial overhead and low efficiency in long-sequence reasoning faced by the Transformer. Moreover, it retains and updates historical information during the encoding and decoding process, enhancing the model's memory and generalization capabilities. Additionally, the Scheduled Sampling method is adopted to alleviate the exposure bias introduced during reasoning, which involves using the model's own output as the input for the next moment with a certain probability during training, instead of using the true label. We conduct the experiment on six public datasets, with the results showing that the proposed model outperforms existing baseline models in terms of precision and generation quality. This paper provides a new solution for the task of sequence-to-sequence multihop knowledge reasoning and also demonstrates the potential application of the Retentive Network in natural language processing. © 2024 IEEE.

关键词： Iterative decoding

来源：评论

学校读者我要写书评

暂无评论

Disentangling Heterogeneous Knowledge Concept Embedding for Cognitive Diagnosis on Untested Knowledge

arXiv

引用

arXiv 2024年

作者： Zhang, Miao Wang, Ziming Xing, Runtian Xiao, Kui Li, Zhifei Zhang, Yan Tang, Chang School of Computer Science and Information Engineering Hubei University China Key Laboratory of Intelligent Sensing System and Security China Hubei Key Laboratory of Big Data Intelligent Analysis and Application China

Cognitive diagnosis is a fundamental and critical task in learning assessment, which aims to infer students’ proficiency on knowledge concepts from their response logs. Current works assume each knowledge concept will certainly be tested and covered by multiple exercises. However, whether online or offline courses, it’s hardly feasible to completely cover all knowledge concepts in several exercises. Restricted tests lead to undiscovered knowledge deficits, especially untested knowledge concepts(UKCs). In this paper, we propose a novel framework for Cognitive Diagnosis called Disentangling Heterogeneous Knowledge Cognitive Diagnosis(DisKCD) on untested knowledge. Specifically, we leverage course grades, exercise questions, and learning resources to learn the potential representations of students, exercises, and knowledge concepts. In particular, knowledge concepts are disentangled into tested and untested based on the limiting actual exercises. We construct a heterogeneous relation graph network via students, exercises, tested knowledge concepts(TKCs), and UKCs. Then, through a hierarchical heterogeneous message-passing mechanism, the fine-grained relations are incorporated into the embeddings of the entities. Finally, the embeddings will be applied to multiple existing cognitive diagnosis models to infer students’ proficiency on UKCs. Experimental results on real-world datasets show that the proposed model can effectively improve the performance of the task of diagnosing students’ proficiency on UKCs. Our code is available at https://***/Hubuers/DisKCD. © 2024, CC BY.

关键词： Students

来源：评论

学校读者我要写书评

暂无评论

Continuous and generalized visual odometry based on FLANN_PSC-RANSAC_LSTM

引用

Journal of intelligent and Fuzzy Systems 2024年第1-2期47卷 15-28页

作者： Li, Dengao Feng, Ran Wu, Fanming Zhao, Jinhua Zhao, Jumin College of Computer Science and Technology College of Data Science Taiyuan University of Technology Shanxi Taiyuan China College of Electronic Information and Optical Engineering Taiyuan University of Technology Shanxi Taiyuan China Key Laboratory of Big Data Fusion Analysis and Application of Shanxi Province Taiyuan University of Technology Shanxi Taiyuan China Intelligent Perception Engineering Technology Center of Shanxi Taiyuan China

In the field of simultaneous localization and mapping (SLAM), visual odometry (VO) always has great application prospects. In recent years, with the progress in the field of machine learning, methods based on neural networks are constantly being updated and applied. In this paper, we propose a continuous and generalized monocular visual odometry method based on features and neural networks. First, the feature information of adjacent image sequences is extracted by matching and troubleshooting algorithm (FLANN_PSC-RANSAC), then it and the corresponding six-degree-of-freedom information are simultaneously input into the long short-term memory artificial neural network (LSTM) for model construction, which not only ensures the reliability of the mode but also eliminates the influence of illumination on the data. In the real environment test, it has been effectively proved in terms of trajectory recovery accuracy and generalization ability to different environments and different illuminations. © 2024 - IOS Press. All rights reserved.

关键词： Long short-term memory

来源：评论

学校读者我要写书评

暂无评论

Temporal Transformation Network Based On Scale Sequences for Cloth-Changing Person Re-Identification in Video datasets 9

Temporal Transformation Network Based On Scale Sequences for...

引用

9th International Conference on Computer and Communications, ICCC 2023

作者： Zhu, Xiaoke Zhao, Bo Dong, Zhiwei Dong, Lingyun Li, Danyang School of Computer and Information Engineering Henan University Kaifeng China Henan University Henan Engineering Research Center of Intelligent Technology and Application China Henan Key Laboratory of Big Data Analysis and Processing Henan University Kaifeng China

ISBN: (纸本)9798350317251

This paper introduces the Temporal Transform Network based on Scale Sequences (TTS) for cloth-changing person re-identification in video datasets. The TTS network is designed to capture multi-scale temporal cues within video sequences. It accomplishes this by initially modeling short-term temporal cues between adjacent frames, followed by capturing long-term relationships between non-consecutive frames. In more detail, short-term temporal cues are modeled through parallel inflated convolutions with different time dilation rates, enabling the representation of pedestrian movement and appearance dynamics. Long-term relationships are effectively captured using a temporal self-attention model, mitigating challenges such as occlusion and noise within the video sequence. The TTS network outperforms existing methods across cloth-changing video ReID datasets such as CCVID. For instance, under general settings, our approach exhibits a 1.1% improvement in top-1 accuracy and a corresponding 1.1% increase in mAP compared to the baseline. In cloth-changing settings, we observe a 0.2% enhancement in top-1 accuracy and a notable 1.3% increase in mAP relative to the baseline. © 2023 IEEE.

关键词： Video recording

来源：评论

学校读者我要写书评

暂无评论

Lightweight Transformer Network and Self-supervised Task for Kinship Verification 8

Lightweight Transformer Network and Self-supervised Task for...

引用

8th IEEE International Conference on Computer and Communications, ICCC 2022

作者： Zhu, Xiaoke Li, Yunwei Li, Danyang Dong, Lingyun Chen, Xiaopan School of Computer and Information Engineering Henan University Kaifeng China Henan Key Laboratory of Big Data Analysis and Processing Henan University Kaifeng China Henan Engineering Research Center of Intelligent Technology and Application Henan University China

ISBN: (纸本)9781665450515

Kinship verification is one of the interesting and critical problems in computer vision research, with significant progress in the past decades. Meanwhile, Vision Transformer (VIT) has recently achieved impressive success in many domains, including object detection, image recognition, and semantic segmentation, among others. Most of the previous work on kinship verification are based on convolutional or recurrent neural networks. Compared with the local processing of images like convolutions, transformers can effectively understand and process images globally. However, due to overuse, there are many Transformer layers of fully connected layers, and VIT speed is still an issue. Therefore, in this paper, inspired by the recent success of Transformer models in vision tasks, we propose a Transformer-based kinship verification for training and optimizing kinship verification models. We first train the basic vision transformer (VIT-B) with 12 transformer layers, then we reduce the transformer layers to 6 layers, namely VIT-S (Small Vit) and 4 layers, namely VIT-T (Tiny Vit), to make a tradeoff between optimization accuracy and efficiency. As the first attempt to apply Transformer to the kinship verification task, it provides a feasible strategy for kinship research topics and verifies the effectiveness of the method in terms of the accuracy of the experimental results. © 2022 IEEE.

关键词： Image recognition

来源：评论

学校读者我要写书评

暂无评论

Point-Level and Set-Level Deep Representation Learning for Cross-Modal Person Re-identification 8

Point-Level and Set-Level Deep Representation Learning for C...

引用

8th IEEE International Conference on Computer and Communications, ICCC 2022

作者： Hu, Jihui Ye, Pengfei Li, Danyang Dong, Lingyun Chen, Xiaopan Zhu, Xiaoke School of Computer and Information Engineering Henan University Kaifeng China Henan University Henan Key Laboratory of Big Data Analysis and Processing Kaifeng China Henan University Henan Engineering Research Center of Intelligent Technology and Application China

ISBN: (纸本)9781665450515

In practice, significant modality differences usually exist between visible and infrared images, which makes visible-infrared Person Re-Identification (VI-ReID) a challenging research task. Due to the existing influence of pose variation, background changes, and occlusion, there are often outlier samples within the set of images from the same person. These outlier samples can adversely affect the process of learning the cross-modal matching model. The existing VI-ReID methods mainly focus on learning cross-modal feature representation by using image-level discriminant constraints, i.e., the distance between the truly-matching cross-modal images should be smaller than that between wrong-matching cross-modal images. However, most of these methods ignore the adverse influence caused by outliers. To solve the above problems, we proposed a Point-level and Set-level Deep Representation Learning (PSDRL) approach for VI-ReID in this paper. By using the set-level constraint in the process of deep representation learning, the discrepancy between visible and infrared modalities can be decreased, and the adverse effect of outliers can be weakened. By employing the image-level constraint, the discriminability of the obtained deep feature representations can be improved. Extensive experiments are conducted on the publicly available cross-modal Person Re-Identification datasets, including SYSU-MM01 and RegDB. Experimental results demonstrate the effectiveness of the proposed approach. © 2022 IEEE.

关键词： Infrared imaging

来源：评论

学校读者我要写书评

暂无评论

Oriented Object Detection in Aerial Images with Double Midpoint Vectors 9

Oriented Object Detection in Aerial Images with Double Midpo...

引用

9th International Conference on Computer and Communications, ICCC 2023

作者： Zhu, Xiaoke Hu, Qi Chen, Xiaopan Yuan, Caihong Li, Minghao Henan University School of Computer and Information Engineering Kaifeng China Henan University Henan Engineering Research Center of Intelligent Technology and Application China Henan University Henan Key Laboratory of Big Data Analysis and Processing Kaifeng China

ISBN: (纸本)9798350317251

In recent years, oriented object detection technology has made significant progress. Most oriented object detectors represent oriented bounding boxes by adding additional angle information to the horizontal bounding box representation, which leads to serious angle periodicity issues. To address these issues, this paper proposes double midpoint vectors (DMvectors) to represent oriented objects. For any oriented bounding box, we align the center point with the origin in the Cartesian coordinate system, and evenly distribute the midpoints of the box edges across the four quadrants. We use the midpoint of the first and second quadrants, as well as the width and height of the box, to represent the oriented bounding box. DMvectors solve the periodic problem of angles by implicitly including angle information in vectors. To demonstrate the effectiveness of our proposed method, we conducted extensive experiments on two common remote sensing datasets (DOTA and HRSC2016), and our method achieved competitive results. © 2023 IEEE.

关键词： Object detection

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：