检索结果-内蒙古大学图书馆

arXiv 2025年

作者： Huang, Junjia Yan, Pengxiang Liu, Jiyang Wu, Jie Wang, Zhao Wang, Yitong Lin, Liang Li, Guanbin Sun Yat-sen University China ByteDance Intelligent Creation China Peng Cheng Laboratory China Guangdong Key Laboratory of Big Data Analysis and Processing China

Image fusion seeks to seamlessly integrate foreground objects with background scenes, producing realistic and harmonious fused images. Unlike existing methods that directly insert objects into the background, adaptive and interactive fusion remains a challenging yet appealing task. It requires the foreground to adjust or interact with the background context, enabling more coherent integration. To address this, we propose an iterative human-in-the-loop data generation pipeline, which leverages limited initial data with diverse textual prompts to generate fusion datasets across various scenarios and interactions, including placement, holding, wearing, and style transfer. Building on this, we introduce DreamFuse, a novel approach based on the Diffusion Transformer (DiT) model, to generate consistent and harmonious fused images with both foreground and background information. DreamFuse employs a Positional Affine mechanism to inject the size and position of the foreground into the background, enabling effective foreground-background interaction through shared attention. Furthermore, we apply Localized Direct Preference Optimization guided by human feedback to refine DreamFuse, enhancing background consistency and foreground harmony. DreamFuse achieves harmonious fusion while generalizing to text-driven attribute editing of the fused results. Experimental results demonstrate that our method outperforms state-of-the-art approaches across multiple metrics. Copyright © 2025, The Authors. All rights reserved.

关键词： Image fusion

来源：评论

学校读者我要写书评

暂无评论

Camera-Conditioned Stable Feature Generation for Isolated Camera Supervised Person Re-IDentification

arXiv

引用

arXiv 2022年

作者： Wu, Chao Ge, Wenhang Wu, Ancong Chang, Xiaobin School of Artificial Intelligence Sun Yat-sen University China Guangdong Key Laboratory of Big Data Analysis and Processing Guangzhou510006 China Key Laboratory of Machine Intelligence and Advanced Computing Ministry of Education China School of Computer Science and Engineering Sun Yat-sen University China

To learn camera-view invariant features for person Re-IDentification (Re-ID), the cross-camera image pairs of each person play an important role. However, such cross-view training samples could be unavailable under the ISolated Camera Supervised (ISCS) setting, e.g., a surveillance system deployed across distant scenes. To handle this challenging problem, a new pipeline is introduced by synthesizing the cross-camera samples in the feature space for model training. Specifically, the feature encoder and generator are end-to-end optimized under a novel method, Camera-Conditioned Stable Feature Generation (CCSFG). Its joint learning procedure raises concern on the stability of generative model training. Therefore, a new feature generator, σ-Regularized Conditional Variational Autoencoder (σReg. CVAE), is proposed with theoretical and experimental analysis on its robustness. Extensive experiments on two ISCS person Re-ID datasets demonstrate the superiority of our CCSFG to the competitors. 1 © 2022, CC BY.

关键词： Cameras

来源：评论

学校读者我要写书评

暂无评论

Dual-Correction Adaptation Network for Noisy Knowledge Transfer

arXiv

引用

arXiv 2022年

作者： Wang, Yunyun Zheng, Weiwen Chen, Songcan The Jiangsu Key Laboratory of Big Data Security & Intelligent Processing Computer Science and Engineering Nanjing University of Posts & Telecommunications Nanjing210046 China The MIIT Key Laboratory of Pattern Analysis and Machine Intelligence Computer Science & Technology/AI Nanjing University of Aeronautics & Astronautics Nanjing210023 China

Previous unsupervised domain adaptation (UDA) methods aim to promote target learning via a single-directional knowledge transfer from label-rich source domain to unlabeled target domain, while its reverse adaption from target to source has not jointly been considered yet so far. In fact, in some real teaching practice, a teacher helps students learn while also gets promotion from students to some extent, which inspires us to explore a dual-directional knowledge transfer between domains, and thus propose a Dual-Correction Adaptation Network (DualCAN) in this paper. However, due to the asymmetrical label knowledge across domains, transfer from unlabeled target to labeled source poses a more difficult challenge than the common source-to-target counterpart. First, the target pseudo-labels predicted by source commonly involve noises due to model bias, hence in the reverse adaptation, they may hurt the source performance and bring a negative target-to-source transfer. Secondly, source domain usually contains innate noises, which will inevitably aggravate the target noises, leading to noise amplification across domains. To this end, we further introduce a Noise Identification and Correction (NIC) module to correct and recycle noises in both domains. To our best knowledge, this is the first naive attempt of dual-directional adaptation for noisy UDA, and naturally applicable to noise-free UDA. A theory justification is given to state the rationality of our intuition. Empirical results confirm the effectiveness of DualCAN with remarkable performance gains over state-of-the-arts, particularly for extreme noisy tasks (e.g., ∼+ 15% on Pw→Pr and Pr→Rw of Office-Home). Copyright © 2022, The Authors. All rights reserved.

关键词： Knowledge management

来源：评论

学校读者我要写书评

暂无评论

Convolutional Long-Short Term Memory Network for convective weather prediction based on 3D Doppler radar data 2020

Convolutional Long-Short Term Memory Network for convective ...

引用

Proceedings of the 2020 International Conference on Cyberspace Innovation of Advanced Technologies

作者： Xiajiong Shen Kunying Meng Zhongli Guo Lei Zhang Henan Key Laboratory of Big Data Analysis and Processing Henan University Kaifeng PR China School of Computer and Information Engineering Henan University Kaifeng PR China Southwest Regional Air Traffic Management Bureau CAAC Senior engineer of aviation meteorology

ISBN: (纸本)9781450387828

Convectional weather is one of the weather which often occurs during the warm season, the scope of this kind of weather is normally small, short duration, etc, so it is difficult to predicted, the effects of convectional weather are very large, from the national economy to military activities will be threatened by it, so how to achieve more accurate of convection weather forecast will be of great significance. Doppler radar is one of the main tools for monitoring and early warning severe convective weather. It can provide important strong convective information for prediction. through its observed real-time echo intensity (reflectivity factor Z), radial velocity (V) and velocity spectrum width (W). Echo intensity is an important basis for judging whether severe, convective weather occurs, and radial velocity can identify wind damage and is also an important basis for judging the occurrence of severe convective weather. From the radial velocity, we can see the convergence, divergence and rotation characteristics of airflow, which are closely related to the occurrence and development of severe convective weather. Therefore, Doppler radar data plays a very important role in the detection and prediction of severe weather. At present, there are still problems about how to effectively extract these information from Doppler radar *** paper proposes a method of near prediction based on deep learning. The neural network is used to automatically learn data features for prediction. Convolution neural network is combined with long-short-time memory network. Firstly, a three-dimensional convolution is proposed to extract the spatial features of three-dimensional original data, and then the long-short-time memory network is used to extract the features of data in time dimension, thus improving the prediction accuracy. Deep learning method avoids the process of manually extracting data features, and uses historical 3D Doppler radar image data to predict convective weather fo

关键词： Convective weather

来源：评论

学校读者我要写书评

暂无评论

Surface soil moisture inversion based on DEFS-GA-BP model using Sentinel-1/2 remote sensing data 5

Surface soil moisture inversion based on DEFS-GA-BP model us...

引用

IET International Radar Conference 2020, IET IRC 2020

作者： Zhang, Chenyang Min, Lin Zhao, Jianhui College of Computer and Information Engineering Henan University Kaifeng475004 China Henan Key Laboratory of Big Data Analysis and Processing Henan University Kaifeng475004 China Henan Engineering Research Center of Intelligent Technology and Application Henan University Kaifeng475004 China Network Information Center Office Henan University Kaifeng475004 China

ISBN: (纸本)9781839535635

The major work is to retrieve the surface soil moisture under crop cover in this paper. This paper proposes an improved GA-BP model called as DEFS-GA-BP model. This model is proposed to retrieve the surface soil moisture distribution using Sentinel-1 and Sentinel-2 remote sensing data. After extracting nine polarization features and one roughness parameter from Sentinel-1 data, five vegetation indexes from Sentinel-2 data. The differential evolution feature selection (DEFS) algorithm is applied to select the optimal feature subset. By using the genetic algorithm (GA) to optimize the node weights, the back propagation (BP) neural network is trained using the optimal feature subset and then used to generate the surface soil moisture distribution. The results of inversion algorithm indicate that the proposed DEFS-GA-BP model has a better inversion precision than GA-BP model, and demonstrates its application potential in surface soil moisture inversion. © 2020 Institution of Engineering and Technology. All rights reserved.

关键词： Genetic algorithms

来源：评论

学校读者我要写书评

暂无评论

Reliability on Deep Learning Models: A Comprehensive Observation

Reliability on Deep Learning Models: A Comprehensive Observa...

引用

International Symposium on System and Software Reliability (ISSSR)

作者： Yuhong Zhang Chunjing Xiao School of Artificial Intelligence and Big Data Key Laboratory of Grain Information Processing and Control Ministry of Education Henan University of Technology Zhengzhou China Henan Key Laboratory of Big Data Analysis and Processing Henan University Kaifeng China

ISBN: (数字)9780738104973

ISBN: (纸本)9781665423236

This paper provides a comprehensive observation to examine the reliability of deep learning (DL) models. First, we will briefly introduce the essential background and kernel techniques in deep learning, such as downsampling and nonlinear discontinuity. Each of them may have some relation to the reliability of DL models. Then we discuss the inherent structural flaws of deep learning and the risk of unreliability that can result from it. Subsequently, we discuss various ways of generating adversarial samples that affect the DL model's reliability and corresponding preventive measures. Finally, we complete this observation by identifying current challenges and future trends for research.

关键词： Reliability Deep learning Biological neural networks Perturbation methods Neural networks Feature extraction data models

来源：评论

学校读者我要写书评

暂无评论

Neural deepfake detection with factual structure of text

Neural deepfake detection with factual structure of text

引用

2020 Conference on Empirical Methods in Natural Language processing, EMNLP 2020

作者： Zhong, Wanjun Tang, Duyu Xu, Zenan Wang, Ruize Duan, Nan Zhou, Ming Wang, Jiahai Yin, Jian The School of Data and Computer Science Sun Yat-sen University Guangdong Key Laboratory of Big Data Analysis and Processing Guangzhou China Microsoft Research Shanghai China Fudan University Shanghai China

ISBN: (纸本)9781952148606

Deepfake detection, the task of automatically discriminating machine-generated text, is increasingly critical with recent advances in natural language generative models. Existing approaches to deepfake detection typically represent documents with coarse-grained representations. However, they struggle to capture factual structures of documents, which is a discriminative factor between machine-generated and human-written text according to our statistical analysis. To address this, we propose a graph-based model that utilizes the factual structure of a document for deepfake detection of text. Our approach represents the factual structure of a given document as an entity graph, which is further utilized to learn sentence representations with a graph neural network. Sentence representations are then composed to a document representation for making predictions, where consistent relations between neighboring sentences are sequentially modeled. Results of experiments on two public deepfake datasets show that our approach significantly improves strong base models built with RoBERTa. Model analysis further indicates that our model can distinguish the difference in the factual structure between machine-generated text and human-written text. © 2020 Association for Computational Linguistics

关键词： Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

Incorporating Domain Knowledge and Semantic Information into Language Models for Commonsense Question Answering

Incorporating Domain Knowledge and Semantic Information into...

引用

International Conference on Computer Supported Cooperative Work in Design

作者： Ruiying Zhou Keke Tian Hanjiang Lai Jian Yin Guangdong Key Laboratory of Big Data Analysis and Processing Guangzhou P.R.China School of Artificial Intelligence Sun Yat-sen University

ISBN: (纸本)9781728165981

Commonsense question answering (CSQA) aims to answer questions which require the system to understand related commonsense knowledge that is not explicitly expressed in the given context. Recent advance in neural language models (e.g., BERT) that are pre-trained on a large-scale text corpus and fine-tuned on downstream tasks has boosted the performance on CSQA. However, due to the lack of domain knowledge (e.g., in social situations), these models fail to reason about specific tasks. In this work, we propose an approach to incorporate domain knowledge and semantic information into language model based approaches for better understanding the related commonsense knowledge. Firstly, we extract the knowledge from existing resources by jointly learning to ask and answer as well as semantic role labeling based answering. These two tasks are correlated and can reinforce each other to discover the domain knowledge. Then, we utilize Semantic Role Labeling to enable the system to gain a better understanding of relations among relevant entities. Experimental results on several CSQA benchmarks demonstrate the effectiveness of the proposed approach.

关键词： Computational modeling Conferences Semantics Bit error rate Benchmark testing Knowledge discovery Collaborative work

来源：评论

学校读者我要写书评

暂无评论

NavCoT: Boosting LLM-Based Vision-and-Language Navigation via Learning Disentangled Reasoning

arXiv

引用

arXiv 2024年

作者： Lin, Bingqian Nie, Yunshuang Wei, Ziming Chen, Jiaqi Ma, Shikui Han, Jianhua Xu, Hang Chang, Xiaojun Liang, Xiaodan Shenzhen Campus of Sun Yat-sen University Shenzhen China Peng Cheng Laboratory Guangdong Key Laboratory of Big Data Analysis and Processing Guangzhou510006 China The University of Hong Kong Hong Kong Dataa Robotics company Huawei Noah’s Ark Lab Hong Kong The School of Information Science and Technology University of Science and Technology of China Hefei230026 China Abu Dhabi United Arab Emirates

Vision-and-Language Navigation (VLN), as a crucial research problem of Embodied AI, requires an embodied agent to navigate through complex 3D environments following natural language instructions. Recent research has highlighted the promising capacity of large language models (LLMs) in VLN by improving navigational reasoning accuracy and interpretability. However, their predominant use in an offline manner usually suffers from substantial domain gap between the VLN task and the LLM training corpus. This paper proposes a novel strategy called Navigational Chain-of-Thought (NavCoT), where we fulfill parameter-efficient in-domain training to enable self-guided navigational decision, leading to a significant mitigation of the domain gap in a cost-effective manner. Specifically, at each timestep, the LLM is prompted to forecast the navigational chain-of-thought by: 1) acting as a world model to imagine the next observation according to the instruction, 2) selecting the candidate observation that best aligns with the imagination, and 3) determining the action based on the reasoning from the prior steps. In this way, the action prediction can be effectively simplified benefiting from the disentangled reasoning. Through constructing formalized labels for training, the LLM can learn to generate desired and reasonable chain-of-thought outputs for improving the action decision. Experimental results across various training settings and popular VLN benchmarks (e.g., Room-to-Room (R2R), Room-across-Room (RxR), Room-for-Room (R4R)) show the significant superiority of NavCoT over the direct action prediction variants. Through simple parameter-efficient finetuning, our NavCoT outperforms a recent GPT4-based approach with ∼7% relative improvement on the R2R dataset. We believe that NavCoT will help unlock more task-adaptive and scalable LLM-based embodied agents, which are helpful for developing real-world robotics applications. Code is available at https://***/expectorlin/NavC

关键词： Robots

来源：评论

学校读者我要写书评

暂无评论

Embedding Dynamic Attributed Networks by Modeling the Evolution Processes 28

Embedding Dynamic Attributed Networks by Modeling the Evolut...

引用

28th International Conference on Computational Linguistics, COLING 2020

作者： Xu, Zenan Ou, Zijing Su, Qinliang Yu, Jianxing Quan, Xiaojun Lin, Zhenkun School of Data and Computer Science Sun Yat-Sen University China Guangdong Key Laboratory of Big Data Analysis and Processing Guangzhou China Key Lab. of Machine Intelligence and Advanced Computing Ministry of Education China

ISBN: (纸本)9781952148279

Network embedding has recently emerged as a promising technique to embed nodes of a network into low-dimensional vectors. While fairly successful, most existing works focus on the embedding techniques for static networks. But in practice, there are many networks that are evolving over time and hence are dynamic, e.g., the social networks. To address this issue, a high-order spatio-temporal embedding model is developed to track the evolutions of dynamic networks. Specifically, an activeness-aware neighborhood embedding method is first proposed to extract the high-order neighborhood information at each given timestamp. Then, an embedding prediction framework is further developed to capture the temporal correlations, in which the attention mechanism is employed instead of recurrent neural networks (RNNs) for its efficiency in computing and flexibility in modeling. Extensive experiments are conducted on four real-world datasets from three different areas. It is shown that the proposed method outperforms all the baselines by a substantial margin for the tasks of dynamic link prediction and node classification, which demonstrates the effectiveness of the proposed methods on tracking the evolutions of dynamic networks. © 2020 COLING 2020 - 28th International Conference on Computational Linguistics, Proceedings of the Conference. All rights reserved.

关键词： Recurrent neural networks

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：