检索结果-内蒙古大学图书馆

Dual Test-time Training for Out-of-distribution Recommender System

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Yang, Xihong Wang, Yiqi Chen, Jin Fan, Wenqi Zhao, Xiangyu Zhu, En Liu, Xinwang Lian, Defu School of Computer National University of Defense Technology Changsha410073 China School of Business and Management Hong Kong University of Science and Technology Hong Kong The Hong Kong Polytechnic University Hong Kong Department of Data Science City University of Hong Kong Hong Kong Anhui Province Key Laboratory of Big Data Analysis and Application School of Computer Science and Technology University of Science and Technology of China Anhui Hefei230000 China

Deep learning has been widely applied in recommender systems, which has recently achieved revolutionary progress. However, most existing learning-based methods assume that the user and item distributions remain unchanged between the training phase and the test phase. However, the distribution of user and item features can naturally shift in real-world scenarios, potentially resulting in a substantial decrease in recommendation performance. This phenomenon can be formulated as an Out-Of-Distribution (OOD) recommendation problem. To address this challenge, we propose a novel Dual Test-Time-Training framework for OOD Recommendation, termed DT3OR. In DT3OR, we incorporate a model adaptation mechanism during the test-time phase to carefully update the recommendation model, allowing the model to adapt specially to the shifting user and item features. To be specific, we propose a self-distillation task and a contrastive task to assist the model learning both the user’s invariant interest preferences and the variant user/item characteristics during the test-time phase, thus facilitating a smooth adaptation to the shifting features. Furthermore, we provide theoretical analysis to support the rationale behind our dual test-time training framework. To the best of our knowledge, this paper is the first work to address OOD recommendation via a test-time-training strategy. We conduct experiments on five datasets with various backbones. Comprehensive experimental results have demonstrated the effectiveness of DT3OR compared to other state-of-the-art baselines. Copyright © 2024, The Authors. All rights reserved.

关键词： Contrastive Learning

Training of Spiking Neural Network joint Curriculum Learning Strategy

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Tang, Lingling Chu, Jielei Gong, Zhiguo Li, Tianrui The School of Computing and Artificial Intelligence Southwest Jiaotong University Chengdu611756 China National Engineering Laboratory of Integrated Transportation Big Data Application Technology Southwest Jiaotong University Chengdu611756 China The State Key Laboratory of Internet of Things for Smart City Department of Computer and Information Science University of Macau China

Starting with small and simple concepts, and gradually introducing complex and difficult concepts is the natural process of human learning. Spiking Neural Networks (SNNs) aim to mimic the way humans process information, but current SNNs models treat all samples equally, which does not align with the principles of human learning and overlooks the biological plausibility of SNNs. To address this, we propose a CL-SNN model that introduces Curriculum Learning(CL) into SNNs, making SNNs learn more like humans and providing higher biological interpretability. CL is a training strategy that advocates presenting easier data to models before gradually introducing more challenging data, mimicking the human learning process. We use a confidence-aware loss to measure and process the samples with different difficulty levels. By learning the confidence of different samples, the model reduces the contribution of difficult samples to parameter optimization automatically. We conducted experiments on static image datasets MNIST, Fashion-MNIST, CIFAR10, and neuromorphic datasets N-MNIST, CIFAR10-DVS, DVS-Gesture. The results are promising. To our best knowledge, this is the first proposal to enhance the biologically plausibility of SNNs by introducing CL. Copyright © 2023, The Authors. All rights reserved.

关键词： Curricula

AdaSemiCD: An Adaptive Semi-supervised Change Detection Method Based on Pseudo Label Evaluation

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Ran, Lingyan Wen, Dongcheng Zhuo, Tao Zhang, Shizhou Zhang, Xiuwei Zhang, Yanning School of Computer Science Shaanxi Provincial Key Laboratory of Speech and Image Information Processing The National Engineering Laboratory for Integrated Aerospace-GroundOcean Big Data Application Technology The Ningbo Institute of Northwestern Polytechnical University Northwestern Polytechnical University Xi’an710072 China College of Information Engineering Northwest A&F University Yangling712100 China

Change detection(CD) is an essential field in remote sensing, with a primary focus on identifying areas of change in bitemporal image pairs captured at varying intervals of the same region. The data annotation process for CD tasks is both time-consuming and labor-intensive. To better utilize the scarce labeled data and abundant unlabeled data, we introduce an adaptive semi-supervised learning method, AdaSemiCD, to improve pseudo-label usage and optimize the training process. Initially, due to the extreme class imbalance inherent in CD, the model is more inclined to focus on the background class, and it is easy to confuse the boundary of the target object. Considering these two points, we develop a measurable evaluation metric for pseudo-labels that enhances the representation of information entropy by class rebalancing and amplification of ambiguous areas, assigning greater weights to prospective change objects. Subsequently, to enhance the reliability of samplewise pseudo-labels, we introduce the AdaFusion module, to dynamically identifying the most uncertain region and substituting it with more trustworthy content. Lastly, to ensure better training stability, we introduce the AdaEMA module, which updates the teacher model using only batches of trusted samples. Experimental results on ten public CD datasets validate the efficacy and generalizability of our proposed adaptive training framework. Copyright © 2024, The Authors. All rights reserved.

关键词： Self-supervised learning

HSR-Diff: Hyperspectral Image Super-Resolution via Conditional Diffusion Models

学校读者我要写书评

暂无评论

HSR-Diff: Hyperspectral Image Super-Resolution via Condition...

International Conference on Computer Vision (ICCV)

作者： Chanyue Wu Dong Wang Yunpeng Bai Hanyu Mao Ying Li Qiang Shen National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology Shaanxi Provincial Key Laboratory of Speech & Image Information Processing School of Computer Science Northwestern Polytechnical University China Department of Computer Science Faculty of Business and Physical Sciences Aberystwyth University U.K.

Despite the proven significance of hyperspectral images (HSIs) in performing various computer vision tasks, its potential is adversely affected by the low-resolution (LR) property in the spatial domain, resulting from multiple physical factors. Inspired by recent advancements in deep generative models, we propose an HSI Super-resolution (SR) approach with Conditional Diffusion Models (HSR-Diff) that merges a high-resolution (HR) multispectral image (MSI) with the corresponding LR-HSI. HSR-Diff generates an HR-HSI via repeated refinement, in which the HR-HSI is initialized with pure Gaussian noise and iteratively refined. At each iteration, the noise is removed with a Conditional Denoising Transformer (CDFormer) that is trained on denoising at different noise levels, conditioned on the hierarchical feature maps of HR-MSI and LR-HSI. In addition, a progressive learning strategy is employed to exploit the global information of full-resolution images. Systematic experiments have been conducted on four public datasets, demonstrating that HSR-Diff outperforms state-of-the-art methods.

关键词：

ACDMSR: Accelerated Conditional Diffusion Models for Single Image Super-Resolution

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Niu, Axi Trung, Pham Xuan Zhang, Kang Sun, Jinqiu Zhu, Yu Kweon, In So Zhang, Yanning School of Computer Science Northwestern Polytechnical University Xi’an710072 China National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology Xi’an710072 China School of Electrical Engineering Korea Advanced Institute of Science and Technology Korea Republic of School of Astronautics Northwestern Polytechnical University Xi’an710072 China

Diffusion models have gained significant popularity in the field of image-to-image translation. Previous efforts applying diffusion models to image super-resolution (SR) have demonstrated that iteratively refining pure Gaussian noise using a U-Net architecture trained on denoising at various noise levels can yield satisfactory high-resolution images from low-resolution inputs. However, this iterative refinement process comes with the drawback of low inference speed, which strongly limits its applications. To speed up inference and further enhance the performance, our research revisits diffusion models in image super-resolution and proposes a straightforward yet significant diffusion model-based super-resolution method called ACDMSR (accelerated conditional diffusion model for image super-resolution). Specifically, our method adapts the standard diffusion model to perform super-resolution through a deterministic iterative denoising process. Our study also highlights the effectiveness of using a pre-trained SR model to provide the conditional image of the given low-resolution (LR) image to achieve superior high-resolution results. We demonstrate that our method surpasses previous attempts in qualitative and quantitative results through extensive experiments conducted on benchmark datasets such as Set5, Set14, Urban100, BSD100, and Manga109. Moreover, our approach generates more visually realistic counterparts for low-resolution images, emphasizing its effectiveness in practical scenarios. © 2023, CC BY.

关键词： Image enhancement

Radiomics based on dual-energy CT for noninvasive prediction of cervical lymph node metastases in patients with nasopharyngeal carcinoma

学校读者我要写书评

暂无评论

Radiography 2025年第4期31卷 102989-102989页

作者： L. Li D. Yang Y. Wu R. Sun Y. Qin M. Kang X. Deng M. Bu Z. Li Z. Zeng X. Zeng M. Jiang B.T. Chen Department of Radiology First Affiliated Hospital of Guangxi Medical University Nanning 530021 Guangxi PR China Department of Radiology Guizhou Provincial People Hospital No.83 East Zhongshan Road Nanming District Guizhou Province 550000 Guiyang PR China Engineering Research Center of Text Computing & Cognitive Intelligence Ministry of Education Key Laboratory of Intelligent Medical Image Analysis and Precise Diagnosis of Guizhou Province State Key Laboratory of Public Big Data College of Computer Science and Technology Guizhou University No. 2870 Huaxi Avenue South Guiyang 550025 Guizhou PR China Department of Nuclear Medicine First Affiliated Hospital of Guangxi Medical University Nanning 530021 Guangxi PR China Department of Radiation Oncology First Affiliated Hospital of Guangxi Medical University Nanning 530021 Guangxi PR China Department of Diagnostic Radiology City of Hope National Medical Center Duarte CA USA

Introduction To develop and validate a machine learning model based on dual-energy computed tomography (DECT) for predicting cervical lymph node metastases (CLNM) in patients diagnosed with nasopharyngeal carcinoma (NPC). Methods This prospective single-center study enrolled patients with NPC and the study assessment included both DECT and 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT). Radiomics features were extracted from each region of interest (ROI) for cervical lymph nodes using arterial and venous phase images at 100 keV and 150 keV, either individually as non-fusion models or combined as fusion models on the DECT images. The performance of the random forest (RF) models, combined with radiomics features, was evaluated by area under the receiver operating characteristic curve (AUC) analysis. DeLong's test was employed to compare model performances, while decision curve analysis (DCA) assessed the clinical utility of the predictive models. Results Sixty-six patients with NPC were included for analysis, which was divided into a training set (n = 42) and a validation set (n = 22). A total of 13 radiomic models were constructed (4 non-fusion models and 9 fusion models). In the non-fusion models, when the threshold value exceeded 0.4, the venous phase at 100 keV (V100) (AUC, 0.9667; 95 % confidence interval [95 % CI], 0.9363–0.9901) model exhibited a higher net benefit than other non-fusion models. The V100 + V150 fusion model achieved the best performance, with an AUC of 0.9697 (95 % CI, 0.9393–0.9907). Conclusion DECT-based radiomics effectively diagnosed CLNM in patients with NPC and may potentially be a valuable tool for clinical decision-making. Implications for practice This study improved pre-operative evaluation, treatment strategy selection, and prognostic evaluation for patients with nasopharyngeal carcinoma by combining DECT and radiomics to predict cervical lymph node status prior to treatment.

关键词： Dual-energy computed tomography (DECT) Nasopharyngeal carcinoma (NPC) Cervical lymph node metastases (CLNM) Radiomics Machine learning model

Spatiotemporal Effects of Built Environment on Varying Rail Transit Station Ridership Patterns

学校读者我要写书评

暂无评论

SSRN

SSRN 2022年

作者： Wang, Jing Wan, Feng Dong, Chunjiao Yin, Chaoying Chen, Xiaoyu Key Laboratory of Transport Industry of Big Data Application Technologies for Comprehensive Transport Ministry of Transport Beijing Jiaotong University Beijing100044 China Hikvision Research Institute Hangzhou310012 China College of Automobile and Traffic Engineering Nanjing Forestry University Nanjing210037 China School of Computer and Information Technology Beijing Key Laboratory of Traffic Data Analysis and Mining Beijing Jiaotong University Beijing100044 China

Although many studies have explored the impacts of the built environment on rail transit station ridership, spatiotemporal heterogeneity of the effects and different ridership patterns are rarely considered. Using multi-source data from China, three direct models (DRMs) based on the geographically and temporally weighted regression (GTWR) model are proposed to reveal the spatiotemporal effects of daily, weekday hourly and weekend hourly ridership patterns. The results show that employment density, entertainment places, housing prices, and entrances/exits are the dominant variables affecting ridership in the three models;however, their effects vary in spatiotemporal ranges. For example, the positive effect of employment density decreases from Monday to Sunday, and its coefficients during the peak hours are highest in the northwest center, where is the largest technology park situated. The results also highlight the importance of bus stations and parking lots in attracting commuter ridership. The findings provide insights into urban rail transit planning. © 2022, The Authors. All rights reserved.

关键词： Housing

Graph Structure Prefix Injection Transformer for Multi-Modal Entity Alignment

学校读者我要写书评

暂无评论

SSRN

SSRN 2024年

作者： Li, Zhifei Luo, Xiangyu Zhang, Miao Zhang, Yan Xiao, Kui School of Computer Science and Information Engineering Hubei University Wuhan430062 China School of Cyber Science and Technology Hubei University Wuhan430062 China Key Laboratory of Intelligent Sensing System and Security Hubei University Ministry of Education Wuhan430062 China Hubei Key Laboratory of Big Data Intelligent Analysis and Application Hubei University Wuhan430062 China

Multi-modal entity alignment aims to integrate equivalent entities in diverse multi-modal knowledge graphs. However, previous studies have primarily focused on multi-modal fusion methods without considering the impact of graph structure heterogeneity on entity alignment tasks due to the distinct structural features of different knowledge graphs. Moreover, the multi-modal information that enriches entity representations, including image, textual content, and numerical attributes, frequently incorporates noise due to modal discrepancies. To tackle these challenges, we introduce the Graph Structural Prefix Injection transformer (GSIEA) for multi-modal entity alignment. This novel approach minimizes structural heterogeneity’s impact and optimizes structural data utilization by integrating a structural prefix inject interaction module within the multi-head attention mechanism. Then, we adopt a convolutional enhancement module to extract local features of multi-modal information and complete feature fusion by cross-modal weights. Extensive experiments on two datasets demonstrate that GSIEA outperforms previous baseline models, showcasing an average performance improvement of 3.26% on MRR, with a maximum enhancement of 12.5%. Additionally, the average performance improvement on Hits@1 is 4.96%, with a maximum enhancement of 16.92%. Our code is available at https://***/HubuKG/GSIEA. © 2024, The Authors. All rights reserved.

关键词： Knowledge graph

Learning from Multi-Perception Features for Real-Word Image Super-resolution

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Niu, Axi Zhang, Kang Pham, Trung X. Wang, Pei Sun, Jinqiu Kweon, In So Zhang, Yanning The School of Computer Science Northwestern Polytechnical University Xi’an710072 China The National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology Xi’an710072 China The School of Electrical Engineering Korea Advanced Institute of Science and Technology Korea Republic of The School of Astronautics Northwestern Polytechnical University Xi’an710072 China

Currently, there are two popular approaches for addressing real-world image super-resolution problems: degradation-estimation-based and blind-based methods. However, degradation-estimation-based methods may be inaccurate in estimating the degradation, making them less applicable to real-world LR images. On the other hand, blind-based methods are often limited by their fixed single perception information, which hinders their ability to handle diverse perceptual characteristics. To overcome this limitation, we propose a novel SR method called MPF-Net that leverages multiple perceptual features of input images. Our method incorporates a Multi-Perception Feature Extraction (MPFE) module to extract diverse perceptual information and a series of newly-designed Cross-Perception Blocks (CPB) to combine this information for effective super-resolution reconstruction. Additionally, we introduce a contrastive regularization term (CR) that improves the model’s learning capability by using newly generated HR and LR images as positive and negative samples for ground truth HR. Experimental results on challenging real-world SR datasets demonstrate that our approach significantly outperforms existing state-of-the-art methods in both qualitative and quantitative measures. © 2023, CC BY.

关键词： Feature extraction