检索结果-内蒙古大学图书馆

Distribution-flexible subset quantization for post-quantizing super-resolution networks

Science China(Information Sciences) 2025年第3期68卷 163-180页

作者： Yunshan ZHONG Mingbao LIN Jingjing XIE Yuxin ZHANG Fei CHAO Rongrong JI Institute of Artificial Intelligence Xiamen University Key Laboratory of Multimedia Trusted Perception and Efficient Computing Ministry of Education of ChinaXiamen University Tencent Youtu Lab Department of Artificial Intelligence School of Informatics Xiamen University Peng Cheng Laboratory

This paper introduces distribution-flexible subset quantization(DFSQ), a post-training quantization method for super-resolution networks. Our motivation for developing DFSQ is based on the distinctive activation distributions of current super-resolution models, which exhibit significant variance across samples and channels. To address this issue, DFSQ conducts channel-wise normalization of the activations and applies distribution-flexible subset quantization(SQ), wherein the quantization points are selected from a universal set consisting of multi-word additive log-scale values. To expedite the selection of quantization points in SQ, we propose a fast quantization points selection strategy that uses K-means clustering to select the quantization points closest to the centroids. Compared to the common iterative exhaustive search algorithm, our strategy avoids the enumeration of all possible combinations in the universal set, reducing the time complexity from exponential to linear. Consequently, the constraint of time costs on the size of the universal set is greatly relaxed. Extensive evaluations of various super-resolution models show that DFSQ effectively improves performance even without fine-tuning. For example, for 4-bit EDSR×2 on the Urban benchmark, DFSQ obtains 0.242 dB PSNR gains.

关键词： super-resolution post-training quantization distribution-flexible subset quantization neural network

来源：评论

学校读者我要写书评

暂无评论

Aligning Text-to-Image Diffusion Models without Human Feedback

Aligning Text-to-Image Diffusion Models without Human Feedba...

引用

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Liu, Tao Kuang, Huafeng Lin, Xianming Key Laboratory of Multimedia Trusted Perception and Efficient Computing Ministry of Education of China Xiamen University 361005 China

ISBN: (纸本)9798350368741

Incorporating human feedback to optimize text-to-image models has demonstrated significant effectiveness. However, the process of collecting high-quality human preference labels is both resource-intensive and time-consuming. To address this challenge, we propose a novel approach that leverages a large language model (LLM) to generate sophisticated prompts, guiding the diffusion model towards enhanced image generation. This process inherently produces ranking pairs that approximate human preferences. We further introduce a novel integration of AI feedback with a Supervised Fine-Tuning (SFT) policy, aligning the model with preference labels derived from AI. Our experiments demonstrate that our approach achieves a notable approximation of human preferences, achieving a performance level of 68.13% compared to human-level benchmarks and delivering competitive results. Furthermore, we showcase the synergistic effects of combining AI feedback with human feedback, resulting in further improvements in image quality. This research offers fresh insights into AI feedback learning within text-to-image generation and lays the groundwork for more efficient and cost-effective training methodologies. © 2025 IEEE.

关键词： AI feedback diffusion model human feedback large language model

来源：评论

学校读者我要写书评

暂无评论

Code Membership Inference for Detecting Unauthorized Data Use in Code Pre-trained Language Models

Code Membership Inference for Detecting Unauthorized Data Us...

引用

2024 Conference on Empirical Methods in Natural Language Processing, EMNLP 2024

作者： Zhang, Sheng Li, Hui Ji, Rongrong Key Laboratory of Multimedia Trusted Perception and Efficient Computing Ministry of Education of China Xiamen University China

ISBN: (纸本)9798891761681

Code pre-trained language models (CPLMs) have received great attention since they can benefit various tasks that facilitate software development and maintenance. However, CPLMs are trained on massive open-source code, raising concerns about potential data infringement. This paper launches the study of detecting unauthorized code use in CPLMs, i.e., Code Membership Inference (CMI) task. We design a framework BUZZER for different settings of CMI. BUZZER deploys several inference techniques, including signal extraction from pre-training tasks, hard-to-learn sample calibration and weighted inference, to identify code membership status accurately. Extensive experiments show that CMI can be achieved with high accuracy using BUZZER. Hence, BUZZER can serve as a CMI tool and help protect intellectual property rights. The implementation of BUZZER is available at: https://***/KDEGroup/Buzzer. © 2024 Association for Computational Linguistics.

关键词： Open source software

来源：评论

学校读者我要写书评

暂无评论

Evaluating and Analyzing Relationship Hallucinations in Large Vision-Language Models 41

Evaluating and Analyzing Relationship Hallucinations in Larg...

引用

41st International Conference on Machine Learning, ICML 2024

作者： Wu, Mingrui Ji, Jiayi Huang, Oucheng Li, Jiale Wu, Yuhang Sun, Xiaoshuai Ji, Rongrong Key Laboratory of Multimedia Trusted Perception and Efficient Computing Ministry of Education of China Xiamen University 361005 China

The issue of hallucinations is a prevalent concern in existing Large Vision-Language Models (LVLMs). Previous efforts have primarily focused on investigating object hallucinations, which can be easily alleviated by introducing object detectors. However, these efforts neglect hallucinations in inter-object relationships, which is essential for visual comprehension. In this work, we introduce R-Bench, a novel benchmark for evaluating Vision Relationship Hallucination. R-Bench features image-level questions that focus on the existence of relationships and instance-level questions that assess local visual comprehension. We identify three types of relationship co-occurrences that lead to hallucinations: relationship-relationship, subject-relationship, and relationship-object. The visual instruction tuning dataset's long-tail distribution significantly impacts LVLMs' understanding of visual relationships. Furthermore, our analysis reveals that current LVLMs tend to disregard visual content and overly rely on the common sense knowledge of Large Language Models. They also struggle with reasoning about spatial relationships based on contextual information. Copyright 2024 by the author(s)

关键词： Visual languages

来源：评论

学校读者我要写书评

暂无评论

StealthDiffusion: Towards Evading Diffusion Forensic Detection through Diffusion Model 24

StealthDiffusion: Towards Evading Diffusion Forensic Detecti...

引用

32nd ACM International Conference on multimedia, MM 2024

作者： Zhou, Ziyin Sun, Ke Chen, Zhongxi Kuang, Huafeng Sun, Xiaoshuai Ji, Rongrong Key Laboratory of Multimedia Trusted Perception and Efficient Computing Ministry of Education of China Xiamen University Fujian Xiamen China

ISBN: (纸本)9798400706868

The rapid progress in generative models has given rise to the critical task of AI-Generated Content Stealth (AIGC-S), which aims to create AI-generated images that can evade both forensic detectors and human inspection. This task is crucial for understanding the vulnerabilities of existing detection methods and developing more robust techniques. However, current adversarial attacks often introduce visible noise, have poor transferability, and fail to address spectral differences between AI-generated and genuine images. To address this, we propose StealthDiffusion, a framework based on stable diffusion that modifies AI-generated images into high-quality, imperceptible adversarial examples capable of evading state-of-the-art forensic detectors. StealthDiffusion comprises two main components: Latent Adversarial Optimization, which generates adversarial perturbations in the latent space of stable diffusion, and Control-VAE, a module that reduces spectral differences between the generated adversarial images and genuine images without affecting the original diffusion model's generation process. Extensive experiments show that StealthDiffusion is effective in both white-box and black-box settings, transforming AI-generated images into high-quality adversarial forgeries with frequency spectra similar to genuine images. These forgeries are classified as genuine by advanced forensic classifiers and are difficult for humans to distinguish. © 2024 ACM.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Face Forgery Detection via Texture and Saliency Enhancement 30th

Face Forgery Detection via Texture and Saliency Enhancemen...

引用

30th International Conference on multimedia Modeling, MMM 2024

作者： Guo, Sizheng Yang, Haozhe Lin, Xianming Key Laboratory of Multimedia Trusted Perception and Efficient Computing Ministry of Education of China Xiamen University Xiamen361005 China

ISBN: (纸本)9783031533044

In recent years, AI-driven advancements have resulted in increasingly sophisticated face forgery techniques, posing a challenge in distinguishing genuine images from manipulated ones. This presents significant societal and trust-related concerns. Most current methods approach this task as a binary classification problem via CNN, which potentially overlooks subtle forgery cues due to the down-sampling operations. This may lead to inadequate extraction of discriminative features. In this paper, we argue that forgery clues are hidden within facial textures and salient regions. Motivated by this insight, we propose the Texture and Saliency Enhancement Network (TSE-Net). TSE-Net contains two primary components: the Dynamic Texture Enhancement Module (DTEM) and the Salient Region Attention Module (SRAM). DTEM employs the Gray-Level Co-occurrence Matrix to extract facial texture information, while SRAM concentrates on salient regions. Extensive experiments demonstrate that the TSE-Net has superior performance in comparison to competing methods, emphasizing its effectiveness in detecting subtle face forgery cues through enhanced texture and saliency information extraction. © 2024, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Textures

来源：评论

学校读者我要写书评

暂无评论

X-Oscar: A Progressive Framework for High-quality Text-guided 3D Animatable Avatar Generation 41

X-Oscar: A Progressive Framework for High-quality Text-guide...

引用

41st International Conference on Machine Learning, ICML 2024

作者： Ma, Yiwei Lin, Zhekai Ji, Jiayi Fan, Yijun Sun, Xiaoshuai Ji, Rongrong Key Laboratory of Multimedia Trusted Perception and Efficient Computing Ministry of Education of China School of Informatics Xiamen University 361005 China

Recent advancements in automatic 3D avatar generation guided by text have made significant progress. However, existing methods have limitations such as oversaturation and low-quality output. To address these challenges, we propose X-Oscar, a progressive framework for generating high-quality animatable avatars from text prompts. It follows a sequential "Geometry→Texture→Animation" paradigm, simplifying optimization through step-by-step generation. To tackle oversaturation, we introduce Adaptive Variational Parameter (AVP), representing avatars as an adaptive distribution during training. Additionally, we present Avatar-aware Score Distillation Sampling (ASDS), a novel technique that incorporates avatar-aware noise into rendered images for improved generation quality during optimization. Extensive evaluations confirm the superiority of X-Oscar over existing text-to-3D and text-to-avatar approaches. Our project page is https://***/Projects/XOscar/. Copyright 2024 by the author(s)

关键词： 3D modeling

来源：评论

学校读者我要写书评

暂无评论

GOI: Find 3D Gaussians of Interest with an Optimizable Open-vocabulary Semantic-space Hyperplane 24

GOI: Find 3D Gaussians of Interest with an Optimizable Open-...

引用

32nd ACM International Conference on multimedia, MM 2024

作者： Qu, Yansong Dai, Shaohui Li, Xinyang Lin, Jianghang Cao, Liujuan Zhang, Shengchuan Ji, Rongrong Fujian Xiamen China Key Laboratory of Multimedia Trusted Perception and Efficient Computing Ministry of Education of China Xiamen University China

ISBN: (纸本)9798400706868

3D open-vocabulary scene understanding, crucial for advancing augmented reality and robotic applications, involves interpreting and locating specific regions within a 3D space as directed by natural language instructions. To this end, we introduce GOI, a framework that integrates semantic features from 2D vision-language foundation models into 3D Gaussian Splatting (3DGS) and identifies 3D Gaussians of Interest using an Optimizable Semantic-space Hyperplane. Our approach includes an efficient compression method that utilizes scene priors to condense noisy high-dimensional semantic features into compact low-dimensional vectors, which are subsequently embedded in 3DGS. During the open-vocabulary querying process, we adopt a distinct approach compared to existing methods, which depend on a manually set fixed empirical threshold to select regions based on their semantic feature distance to the query text embedding. This traditional approach often lacks universal accuracy, leading to challenges in precisely identifying specific target areas. Instead, our method treats the feature selection process as a hyperplane division within the feature space, retaining only those features that are highly relevant to the query. We leverage off-the-shelf 2D Referring Expression Segmentation (RES) models to fine-tune the semantic-space hyperplane, enabling a more precise distinction between target regions and others. This fine-tuning substantially improves the accuracy of open-vocabulary queries, ensuring the precise localization of pertinent 3D Gaussians. Extensive experiments demonstrate GOI's superiority over previous state-of-the-art methods. The dataset, model, and code are available at https://***/GOI-Hyperplane/. © 2024 ACM.

关键词： Augmented reality

来源：评论

学校读者我要写书评

暂无评论

DiagnoCare: A Diagnostic Analysis System for Electronic Medical Records Combining Medical Knowledge Graph and Large Language Model 10

DiagnoCare: A Diagnostic Analysis System for Electronic Medi...

引用

10th IEEE Smart World Congress, SWC 2024

作者： Huang, Shiwang Luo, Guofeng Wang, Jiaru Chen, Longbiao Xiamen University Fujian Key Laboratory of Sensing and Computing for Smart Cities China Xiamen University Key Laboratory of Multimedia Trusted Perception and Efficient Computing Xiamen China

ISBN: (纸本)9798331520861

With the rapid development of healthcare technologies and the increasing demand for intelligent diagnostic systems, we present a novel approach that combines medical knowledge graphs and Large Language Models (LLMs) for enhanced Electronic Medical Record (EMR) analysis and condition warning. Our system integrates structured medical knowledge with advanced text analysis capabilities to accurately interpret and process complex EMR data. By leveraging LLMs and prompt learning, the system provides personalized examination suggestions, diagnostic analyses, and disease warnings. Experimental results on real-world datasets, including the MIMIC III database and anonymized hospital records from Fujian Province, China, demonstrate the system's efficacy in delivering precise and actionable insights. Case studies further illustrate the system's practical value and usability, highlighting its potential to significantly improve clinical decision-making and patient outcomes. This research offers substantial implications for the advancement of intelligent EMR systems and personalized healthcare. © 2024 IEEE.

关键词： Electronic health record

来源：评论

学校读者我要写书评

暂无评论

The Dormant Neuron Phenomenon in Multi-Agent Reinforcement Learning Value Factorization 38

The Dormant Neuron Phenomenon in Multi-Agent Reinforcement L...

引用

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Qin, Haoyuan Ma, Chennan Deng, Mian Liu, Zhengzhu Mei, Songzhu Liu, Xinwang Wang, Cheng Shen, Siqi China Key Laboratory of Multimedia Trusted Perception and Efficient Computing XMU China School of Computer National University of Defense Technology China

In this work, we study the dormant neuron phenomenon in multi-agent reinforcement learning value factorization, where the mixing network suffers from reduced network expressivity caused by an increasing number of inactive neurons. We demonstrate the presence of the dormant neuron phenomenon across multiple environments and algorithms, and show that this phenomenon negatively affects the learning process. We show that dormant neurons correlates with the existence of over-active neurons, which have large activation scores. To address the dormant neuron issue, we propose ReBorn, a simple but effective method that transfers the weights from over-active neurons to dormant neurons. We theoretically show that this method can ensure the learned action preferences are not forgotten after the weight-transferring procedure, which increases learning effectiveness. Our extensive experiments reveal that ReBorn achieves promising results across various environments and improves the performance of multiple popular value factorization approaches. The source code of ReBorn is available in https://***/xmu-rl-3dv/ReBorn. © 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：