检索结果-内蒙古大学图书馆

6th IFIP TC 12 International Conference on Intelligence science, ICIS 2024

作者： Li, Dingyi PCA Lab Key Lab of Intelligent Perception and Systems for High-Dimensional Information of Ministry of Education and Jiangsu Key Lab of Image and Video Understanding for Social Security School of Computer Science and Engineering Nanjing University of Science and Technology Nanjing210094 China

ISBN: (纸本)9783031712524

Perceptual video super-resolution aims at converting low-resolution videos to visually appealing high-resolution ones. It may lead to temporal inconsistency due to the drastically changing outputs. In this paper, we propose cascaded sliding-window-based relativistic GAN (Generative Adversarial Network) fusion for perceptual and consistent video super-resolution (PC-VSR). Firstly, cascaded sliding-window-based relativistic GAN is designed to extract more useful information. It enlarges the temporal receptive field of sliding-window-based model in each step. It is able to enhance perceptual quality and compensate temporal consistency progressively and sufficiently. The trained separate refinement generator networks are fused into a final refinement generator. The final refinement generator can be calculated recursively at the testing stage. With our generator fusion, the parameter number is reduced and good quality is maintained. Extensive experimental results demonstrate that our approach outperforms state-of-the-art super-resolution methods in terms of perceptual quality. Our method also achieves good temporal consistency and per-pixel accuracy, compared with other perceptual approaches. © IFIP International Federation for information processing 2025.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Leveraging Timbre Semantic Descriptors for Cross-Dataset Instrument Recognition

引用

IEEE Transactions on Audio, Speech and Language processing 2025年 33卷 1196-1207页

作者： Shenyang Xu Yuan Wang Zijin Li Feng Yu Wei Li Department of Music AI and Information Technology Central Conservatory of Music Beijing China School of Computer Science and Technology Fudan University Shanghai China Shanghai Key Laboratory of Intelligent Information Processing Fudan University Shanghai China

Timbre, a perceptual attribute of musical sound, plays a critical role in both instrument recognition and timbre perception research. However, these fields are rarely integrated, and as a result, their findings rarely benefit one another. This study investigates how integrating timbre semantic descriptors as an auxiliary task can enhance musical instrument recognition under challenging cross-dataset conditions via multi-task learning (MTL). A standard flat MTL framework is introduced, alongside an asymmetric MTL proposed to amplify the influence of timbre semantic descriptors on instrument recognition. Experiments are conducted on Western and Chinese instrument datasets to evaluate cross-cultural performance. The results demonstrate significant improvements on both datasets using the flat MTL, with further analysis revealing the varying contributions of individual timbre descriptors and demonstrating its robustness across different hyperparameter settings. The asymmetric MTL achieves additional performance gains on Western datasets but not on Chinese ones. A dedicated analysis is carried out to investigate this discrepancy, offering insights into potential dataset-specific or cultural challenges.

关键词： Instruments Timbre Semantics Vectors Training Accuracy Testing Multitasking Standards Speech processing

来源：评论

学校读者我要写书评

暂无评论

Corrigendum to "Hyperledger sawtooth based supplychain traceability system for counterfeit drugs" [Comput. Ind. Eng. 190 (2024) 110021] (computers & Industrial Engineering (2024) 190, (S0360835224001426), (10.1016/***.2024.110021))

引用

computers and Industrial Engineering 2025年 204卷

作者： Nawaz, Anum Wang, Liguan Irfan, Muhammad Westerlund, Tomi Shanghai Key Laboratory of Intelligent Information Processing School of Computer Science Fudan University Shanghai China Turku Intelligent and Embedded Robotic Systems Lab Faculty of Technology University of Turku Finland

关键词：

来源：评论

学校读者我要写书评

暂无评论

EVE: Towards End-to-End Video Subtitle Extraction with Vision-Language Models

arXiv

引用

arXiv 2025年

作者： Yu, Haiyang Lu, Jinghui Wang, Yanjie Li, Yang Wang, Han Huang, Can Li, Bin Shanghai Key Laboratory of Intelligent Information Processing China School of Computer Science Fudan University China ByteDance Inc China

The advent of Large Vision-Language Models (LVLMs) has advanced the video-based tasks, such as video captioning and video understanding. Some previous research indicates that taking texts in videos as input can further improve the performance of video understanding. As a type of indispensable information in short videos or movies, subtitles can assist LVLMs to better understand videos. Most existing methods for video subtitle extraction are based on a multi-stage framework, handling each frame independently. They can hardly exploit the temporal information of videos. Although some LVLMs exhibit the robust OCR capability, predicting accurate timestamps for subtitle texts is still challenging. In this paper, we propose an End-to-end Video Subtitle Extraction method, called EVE, which consists of three modules: a vision encoder, an adapter module, and a large language model. To effectively compress the visual tokens from the vision encoder, we propose a novel adapter InterleavedVT to interleave two modalities. It contains a visual compressor and a textual region compressor. The proposed InterleavedVT exploits both the merits of average pooling and Q-Former in token compression. Taking the temporal information of videos into account, we introduce a sliding-window mechanism in the textual region compressor. To benchmark the video subtitle extraction task, we propose a large dataset ViSa including 2.5M videos. Extensive experiments on ViSa demonstrate that the proposed EVE can outperform existing open-sourced tools and LVLMs. Copyright © 2025, The Authors. All rights reserved.

关键词： Visual languages

来源：评论

学校读者我要写书评

暂无评论

Towards Multi-label Open Set Diagnosis for Medical Images

Towards Multi-Label Open Set Diagnosis for Medical Images

引用

IEEE International Symposium on Biomedical Imaging

作者： Shijie Pang Yuze Zhao Rui-Wei Zhao Rui Feng Academy for Engineering and Technology Fudan University Shanghai China Shanghai Key Laboratory of Intelligent Information Processing School of Computer Science Fudan University Shanghai China

ISBN: (数字)9798331520526

ISBN: (纸本)9798331520533

In real clinical scenarios, diagnosis algorithms may face the complex concurrent presence of multiple common, rare, or even unknown disease conditions. In particular, rare disease categories may be underrepresented or even absent from the class-imbalanced training set, leaving safe and trustworthy diagnostic decision-making extremely challenging. In this work, we first formalize the medical image diagnosis as a multi-label open set recognition under class-imbalanced training data and propose a novel Multi-label Open set Diagnosis (MOSD) solution to address this problem. Our MOSD first employs a Balanced Multi-label Data Augmentation (BMDA) module to augment multi-label image data to alleviate the class imbalance in model training more effectively. Then it also boasts a Multi-label Evidential Deep Learning (MEDL) module with tailored learning objectives to better estimate the uncertainty for multi-label open set prediction. Experimental results on the ODIR-5K benchmark successfully demonstrate its superiority in both close set known class classification and open set unknown class detection.

关键词： Training Uncertainty Image recognition Face recognition Training data Multi label classification Robustness Medical diagnosis Medical diagnostic imaging Diseases

来源：评论

学校读者我要写书评

暂无评论

RBDQ: A Reliable LLM-based Text-to-SQL System for Business Data Queries 25

RBDQ: A Reliable LLM-based Text-to-SQL System for Business D...

引用

Companion Proceedings of the ACM on Web Conference 2025

作者： Fenglin Bi Dongdong Cao Zhiyu Wang Yang Chen Fangliang Zhao Tao Hu Zhi Li Yanbin Zhang Wei Wang School of Data Science and Engineering East China Normal University Shanghai China ByteDance Inc Shanghai China School of Computer Science Fudan University Shanghai China and Shanghai Key Lab of Intelligent Information Processing Shanghai China

ISBN: (纸本)9798400713316

Using large language models (LLMs) to convert natural language (NL) into SQL simplifies data access for users by allowing them to use everyday language. However, business departments often distrust LLM-based text-to-SQL systems due to the probabilistic nature of SQL generation, which can result in incorrect but executable SQL queries caused by model hallucinations. This leads to significant concerns regarding the accuracy and reliability of the queried data. In this paper, we present RBDQ, a novel LLM-based text-to-SQL system designed to address the unique challenges of business data queries. RBDQ innovatively introduces the Hierarchical Metrics Query Method and integrates advanced Retrieval-Augmented Generation (RAG) methods along with a self-reflection mechanism to tackle these challenges. RBDQ effectively meets the requirements of business metric queries in real-world scenarios. Currently implemented in the Quality Assurance department at ByteDance, RBDQ has significantly improved operational efficiency and query flexibility. Our experiments demonstrate the system's effectiveness, achieving an Execution Accuracy of 96.20%.

关键词： business data queries

来源：评论

学校读者我要写书评

暂无评论

Efficient Image Super-Resolution With Feature Interaction Weighted Hybrid Network

引用

IEEE Transactions on Multimedia 2025年 27卷 2256-2267页

作者： Li, Wenjie Li, Juncheng Gao, Guangwei Deng, Weihong Yang, Jian Qi, Guo-Jun Lin, Chia-Wen Beijing University of Posts and Telecommunications Pattern Recognition and Intelligent System Laboratory School of Artificial Intelligence Beijing100080 China Shanghai University School of Communication and Information Engineering Shanghai200444 China Nanjing University of Posts and Telecommunications IVIPLab Institute of Advanced Technology Nanjing210046 China Ministry of Education Key Laboratory of Artificial Intelligence Shanghai200240 China Soochow University Provincial Key Laboratory for Computer Information Processing Technology Suzhou215006 China Nanjing University of Science and Technology School of Computer Science and Technology Nanjing210094 China Westlake University Research Center for Industries of the Future School of Engineering Hangzhou310024 China OPPO Research SeattleWA98101 United States National Tsing Hua University Department of Electrical Engineering Institute of Communications Engneering Hsinchu300044 Taiwan

Lightweight image super-resolution aims to reconstruct high-resolution images from low-resolution images using low computational costs. However, existing methods result in the loss of middle-layer features due to activation functions. To minimize the impact of intermediate feature loss on reconstruction quality, we propose a Feature Interaction Weighted Hybrid Network (FIWHN), which comprises a series of Wide-residual Distillation Interaction Block (WDIB) as the backbone. Every third WDIB forms a Feature Shuffle Weighted Group (FSWG) by applying mutual information shuffle and fusion. Moreover, to mitigate the negative effects of intermediate feature loss, we introduce Wide Residual Weighting units within WDIB. These units effectively fuse features of varying levels of detail through a Wide-residual Distillation Connection (WRDC) and a Self-Calibrating Fusion (SCF). To compensate for global feature deficiencies, we incorporate a Transformer and explore a novel architecture to combine CNN and Transformer. We show that our FIWHN achieves a favorable balance between performance and efficiency through extensive experiments on low-level and high-level tasks. © 1999-2012 IEEE.

关键词： Transformers Convolutional neural networks Computational modeling Feature extraction Training Superresolution Adaptation models computer architecture Telecommunications Image reconstruction

来源：评论

学校读者我要写书评

暂无评论

FediLive: A Framework for Collecting and Preprocessing Snapshots of Decentralized Online Social Networks 25

FediLive: A Framework for Collecting and Preprocessing Snaps...

引用

Companion Proceedings of the ACM on Web Conference 2025

作者： Shaojie Min Shaobin Wang Yaxiao Luo Min Gao Qingyuan Gong Yu Xiao Yang Chen Shanghai Key Lab of Intelligent Information Processing School of Computer Science Fudan University Shanghai China Research Institute of Intelligent Complex Systems Fudan University Shanghai China Department of Information and Communications Engineering Aalto University Espoo Finland

ISBN: (纸本)9798400713316

Decentralized online social networks such as Mastodon have emerged quickly during the past years, and offer unique opportunities to investigate user behavior, moderation strategies, and community evolution. However, their decentralized nature imposes challenges for data collection and preprocessing, particularly for obtaining a real-time snapshot in a timely manner. This paper introduces FediLive, a framework designed to rapidly collect and preprocess the live feeds from Mastodon, generate a comprehensive snapshot in real-time, including user-generated contents, interaction networks, and users' demographic attributes. Such a snapshot could further be leveraged for data analysis from different angles, leading to a deeper understanding of user activities on Mastodon. Using FediLive, we collected a 13-day snapshot of Mastodon, covering the publicly-visible activities of all Mastodon users. Our study demonstrates the usefulness of FediLive, and reveals its potential in facilitating data-driven analysis for decentralized online social networks.

关键词： decentralized online social networks

来源：评论

学校读者我要写书评

暂无评论

Distributed Multi-Agent Resource Allocation Based on Transformer and DRL for Cloud XR Hybrid Content Transmission

引用

IEEE Internet of Things Journal 2025年

作者： Wang, Zhaocheng Wu, Jun Wang, Rui Li, Ying Tongji University School of Computer Science and Technology Shanghai201804 China Fudan University School of Computer Science Shanghai Key Laboratory of Intelligent Information Processing Shanghai200433 China Tongji University College of Electronics and Information Engineering Shanghai201804 China Tongji University Shanghai Institute of Intelligent Science and Technology Shanghai201804 China Peng Cheng Laboratory Guangdong518055 China

Extended reality (XR) technologies and applications have grown rapidly in recent years. In addition to providing immersive ultra-high-definition (UHD) video, XR also allows for a haptic experience where devices can be remotely manipulated to accomplish tasks. However, varying numbers of XR users accessing the communication system can strain limited spectrum resources, posing challenges in resource allocation. Therefore, this paper studies resource blocks (RBs) allocation problem in a downlink transmission scenario where real-time cloud XR video and haptic contents need to be transmitted simultaneously. We also consider the random variation in the number of XR users and propose an Adaptive Distributed Multi-Agent Deep reinforcement learning (DRL) combined with Transformer (ADMA-DcT) for dynamic RB allocation method. This method addresses the dynamic change in network input dimensions due to user number variability using a state division module and a self-attention mechanism in the encoder module. To our knowledge, this is the first work to study the RB allocation problem in Cloud XR transmission considering simultaneous transmitting of video and haptic services with a dynamically changing user base. Our extensive simulations show that the ADMA-DcT model, end-to-end trained, outperforms other benchmarks in successfully serving a larger number of XR users under varying user number conditions, demonstrating excellent adaptivity and robustness. © 2014 IEEE.

关键词： Multi agent systems

来源：评论

学校读者我要写书评

暂无评论

DeTrack: In-model Latent Denoising Learning for Visual Object Tracking

arXiv

引用

arXiv 2025年

作者： Zhou, Xinyu Li, Jinglun Hong, Lingyi Jiang, Kaixun Guo, Pinxue Ge, Weifeng Zhang, Wenqiang Shanghai Key Lab of Intelligent Information Processing School of Computer Science Fudan University Shanghai China Shanghai Engineering Research Center of AI & Robotics Academy for Engineering and Technology Fudan University Shanghai China

Previous visual object tracking methods employ image-feature regression models or coordinate autoregression models for bounding box prediction. Image-feature regression methods heavily depend on matching results and do not utilize positional prior, while the autoregressive approach can only be trained using bounding boxes available in the training set, potentially resulting in suboptimal performance during testing with unseen data. Inspired by the diffusion model, denoising learning enhances the model's robustness to unseen data. Therefore, We introduce noise to bounding boxes, generating noisy boxes for training, thus enhancing model robustness on testing data. We propose a new paradigm to formulate the visual object tracking problem as a denoising learning process. However, tracking algorithms are usually asked to run in real-time, directly applying the diffusion model to object tracking would severely impair tracking speed. Therefore, we decompose the denoising learning process into every denoising block within a model, not by running the model multiple times, and thus we summarize the proposed paradigm as an in-model latent denoising learning process. Specifically, we propose a denoising Vision Transformer (ViT), which is composed of multiple denoising blocks. In the denoising block, template and search embeddings are projected into every denoising block as conditions. A denoising block is responsible for removing the noise in a predicted bounding box, and multiple stacked denoising blocks cooperate to accomplish the whole denoising process. Subsequently, we utilize image features and trajectory information to refine the denoised bounding box. Besides, we also utilize trajectory memory and visual memory to improve tracking stability. Experimental results validate the effectiveness of our approach, achieving competitive performance on several challenging datasets. The proposed in-model latent denoising tracker achieve real-time speed, rendering denoising learning

关键词： Regression analysis

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：