检索结果-内蒙古大学图书馆

An infrastructure software perspective toward computation offloading between executable specifications and foundation models

学校读者我要写书评

暂无评论

science China(Information sciences) 2025年第4期68卷 380-382页

作者： Dezhi RAN Mengzhou WU Yuan CAO Assaf MARRON David HAREL Tao XIE Key Laboratory of High Confidence Software Technologies (PKU) Ministry of Education School of Computer SciencePeking University School of Electronics Engineering and Computer Science Peking University Department of Computer Science and Applied Mathematics Weizmann Institute of Science

Foundation models(FMs) [1] have revolutionized software development and become the core components of large software systems. This paradigm shift, however, demands fundamental re-imagining of software engineering theories and methodologies [2]. Instead of replacing existing software modules implemented by symbolic logic, incorporating FMs' capabilities to build software systems requires entirely new modules that leverage the unique capabilities of ***, while FMs excel at handling uncertainty, recognizing patterns, and processing unstructured data, we need new engineering theories that support the paradigm shift from explicitly programming and maintaining user-defined symbolic logic to creating rich, expressive requirements that FMs can accurately perceive and implement.

关键词：

Underwater Biological Target Detection Algorithm and Research Based on YOLOv7 Algorithm

学校读者我要写书评

暂无评论

IAENG International Journal of computer science 2024年第6期51卷 594-601页

作者： Zhuang, Hongwei Liu, Weisheng School of Computer Science and Software Engineering University of Science and Technology Liaoning Anshan China College of Computer Science and Software Engineering University of Science and Technology Liaoning CO Anshan114051 China

Underwater target detection is an important method for detecting marine organisms. However, due to the image occlusion of underwater targets, blurred water quality, poor lighting conditions, small targets, and complex backgrounds, the detection of underwater biological targets has posed significant challenges. In the intricate underwater environment, the conventional feature extraction method has a few drawbacks, including imprecise feature extraction, sluggish detection speed, and inadequate robustness. Consequently, an underwater target detection method based on the enhanced You Only Look Once 7 (YOLOv7) is proposed in this study. The network architecture is reconstructed, and the Deformable Convolutional Network (DCN) modules replace some 3×3 convolutional blocks in the ELAN structure to offset sampling points and reduce background interference. Skip connections and 1× 1 convolutional architecture are added to the DCN module to improve the model’s perception of image details. In addition, Contextual Transformer 3 (COT3) is also incorporated to improve visual performance. Finally, to improve the detection efficiency of small objects, the CIoU loss function is finally replaced by the Normalized Wasserstein Distance (NWD) algorithm. The mAP of DCCN-YOLOv7 on the URPC dataset is 80.4%, according to the experimental results, 2.8% higher than the YOLOv7 network model that is used as a baseline. Furthermore, in contrast to the original YOLOv7 algorithm, the detection speed and accuracy are higher, making it more appropriate for target recognition underwater. © (2024), (International Association of Engineers). All rights reserved.

关键词： Feature extraction

MindScore: quantifying human preference for text-to-image generation through multi-view lens

学校读者我要写书评

暂无评论

science China(Information sciences) 2025年第6期68卷 72-85页

作者： Yiqi TONG Jiarui ZHANG Shaohang WEI Wei GUO Fuzhen ZHUANG Deqing WANG Xi YANG Richeng XUAN School of Artificial Intelligence Beihang University School of Computer Science and Engineering Beihang University Department of Computer Science and Engineering Shanghai Jiao Tong University School of Computer Science Peking University State Key Laboratory of Complex & Critical Software Environment Beihang University Beijing Academy of Artificial Intelligence

Understanding and quantifying the capabilities of foundation models, particularly in text-to-image(T2I) generation, is crucial for verifying their alignment with human expectations and practical requirements. However, evaluating T2I foundation models presents significant challenges due to the complex, multi-dimensional psychological factors that influence human preferences for generated images. In this work, we propose MindScore, a multi-view framework for assessing the generation capacity of T2I models through the lens of human preference. Specifically, MindScore decomposes the evaluation into four complementary modules that align with human cognitive processing of images: matching, faithfulness, quality,and realness. The matching module quantifies the semantic alignment between generated images and prompt text, while the faithfulness module measures how accurately the images reflect specific prompt details. Furthermore, we incorporate quality and realness modules to capture deeper psychological preferences, recognizing that unpleasant or distorted images often trigger adverse human responses. Extensive experiments on three T2I datasets with human preference annotations clearly validate the superiority of our proposed MindScore over various state-of-the-art baselines. Our case studies further reveal that MindScore offers valuable insights into T2I generation from a human-centric perspective.

关键词： text-to-image generation foundation models human preference evaluation multi-view assessment language and vision

Object Detection Model for Remote Sensing Images Based on YOLOv9

学校读者我要写书评

暂无评论

IAENG International Journal of computer science 2025年第3期52卷 840-847页

作者： Hou, Donghao Zhang, Yujun School of Computer and Software Engineering University of Science and Technology Liaoning Anshan114051 China

In the field of object detection for remote sensing images, especially in applications such as environmental monitoring and urban planning, significant progress has been made. This paper addresses the common challenges faced by traditional object detection methods in remote sensing images, such as the large number of targets and complex backgrounds, by proposing a novel network based on YOLOv9. The network innovatively introduces the C3_CD_CGA module, an enhanced module based on Cascaded Group Attention, designed to reduce computational redundancy and increase attention diversity, and enhances the processing capability of multi-scale information through the CD module. The C3 module employs deep asymmetric convolution to mitigate information loss and increase the receptive field. Additionally, the network integrates DSConv with the RepNCSPELAN4 module to adaptively focus on and precisely capture the features of elongated and curved local structures, such as vehicles. The introduction of the CARAFE module further improves the spatial resolution of the feature maps, significantly enhancing performance across various visual tasks. Experimental results show that the improved YOLOv9 achieves a mean average precision (mAP) of 88% on the SIMD dataset, which is an improvement of 1.6% compared to the baseline YOLOv9 model and 1.5% higher than the state-of-the-art YOLO-SE model. This model not only achieves more effective multi-target recognition in complex backgrounds but also strikes a good balance between accuracy and efficiency. © (2025), (International Association of Engineers). All rights reserved.

关键词： Urban planning

A Transfer Learning Framework for Deep Multi-Agent Reinforcement Learning

学校读者我要写书评

暂无评论

IEEE/CAA Journal of Automatica Sinica 2024年第11期11卷 2346-2348页

作者： Yi Liu Xiang Wu Yuming Bo Jiacun Wang Lifeng Ma the School of Automation Nanjing University of Science and Technology the Department of Computer Science and Software Engineering Monmouth University

Dear Editor,This letter presents a new transfer learning framework for the deep multi-agent reinforcement learning(DMARL) to reduce the convergence difficulty and training time when applying DMARL to a new scenario [1... 详细信息

关键词： Deep agent Framework

Research on Image Defogging Algorithm Based on Improved FFA-Net

学校读者我要写书评

暂无评论

IAENG International Journal of computer science 2024年第6期51卷 634-641页

作者： Qinrong, Li Chi, Ma Qiang, Guo Hui, Hu School of Computer Science and Software Engineering University of Science and Technology LiaoNing AnShan114051 China School of Computer Science and Engineering Huizhou University Huizhou516007 China

Images captured under severe weather conditions, such as haze and fog, suffer from image quality degradation caused by atmospheric particle diffusion. This degradation manifests as color fading, reduced contrast, and adversely affects the performance of various computer vision tasks. To address this, this paper presents an end-to-end feature fusion attention network (FFA-Net) designed to directly restore haze-free images. By incorporating the SSIM loss into the original loss function, the proposed method effectively captures the visual disparities between the estimated defogged image and the authentic haze-free image. Additionally, it mitigates the color distortion problem inherent in the original algorithm. To address the challenge of low brightness in input images, a low illumination enhancement module is introduced, seamlessly integrated with the FFA-Net defogging method. Subsequently, a comparative analysis of different defogging algorithms is conducted using two distinct foggy datasets. Multiple evaluation metrics are employed to assess the performance of these algorithms. The findings indicate that our algorithm significantly outperforms others in terms of objective indicators such as PSNR and SSIM, as well as visual effects. © (2024), (International Association of Engineers). All rights reserved.

关键词： Image enhancement

An Apricot Detection Algorithm in Complex Environments Based on Improved YOLOv7

学校读者我要写书评

暂无评论

IAENG International Journal of computer science 2024年第12期51卷 2135-2144页

作者： Guo, Qiang Ma, Chi Hu, Hui School of Computer Science and Software Engineering University of Science and Technology LiaoNing AnShan114051 China School of Computer Science and Engineering Huizhou University Huizhou516007 China

Apricot detection is a prerequisite for counting and harvesting tasks. Existing algorithms face challenges in adapting to the impacts of complex environmental factors such as lighting variations, shadows, dense foliage, and the uneven distribution of samples in mechanized apricot harvesting. This paper proposes an enhanced model, YOLOv7-DC, based on YOLOv7, to address these challenges. YOLOv7-DC preprocesses diverse apricot tree samples to accommodate real-world harvesting detection scenarios. To improve model inference speed and detection accuracy, the detection network is redesigned with a new feature fusion method. DCNv2 is embedded within the efficient layer aggregation network (ELAN), and PConv is introduced to replace conventional convolutions, reducing the parameter impact of DCNv2. The training process incorporates the CBAM attention mechanism to enhance spatial and channel information. The ConvMixer architecture captures spatial and channel relationships transmitted to the detection head through the attention mechanism, improving the model’s detection accuracy for each specific classification sample. Experimental results show that YOLOv7-DC maintains good detection speed and recognition rates across various classification tasks. The improved model achieves a 6.2% increase in average detection accuracy compared to previous algorithms, with a 13% reduction in model parameters. YOLOv7-DC is better suited for handling imbalanced samples and complex environmental scenarios. © (2024), (International Association of Engineers). All rights reserved.

关键词： Apricot biloba detection Attention mechanism Feature fusion YOLOv7

Improved Road Damage Detection Algorithm Based on YOLOv8n

学校读者我要写书评

暂无评论

IAENG International Journal of computer science 2024年第11期51卷 1720-1730页

作者： Li, Xudong Zhang, Yujun School of Computer Science and Software Engineering University of Science and Technology Liaoning Anshan114051 China

This paper introduces an advanced road damage detection algorithm that effectively addresses the shortcomings of existing models, including limited detection performance and large parameter sizes, by utilizing the YOLOv8n model. Key enhancements are integrated into the proposed algorithm to bolster its efficacy. First, the ConvNeXt V2 backbone network is integrated to improve the extraction of contextual features, thereby enhancing the effectiveness of road damage detection. Second, the algorithm employs a C2f_GhostNetV2 block structure to strengthen feature representation while simultaneously reducing computational costs. Additionally, PConv is utilized in the neck region to optimize spatial feature extraction, thereby minimizing redundant computations. The experimental results indicate that the proposed algorithm performs effectively on the Chinese subset of the RDD2022 dataset. Specifically, detection accuracy improved by 1.9%, recall by 1.8%, and mAP@0.5 by 2.1%, while the number of parameters decreased by 24% compared to the initial model. The optimized algorithm increases FPS by 4, meeting the dual requirements of mobile devices for accuracy and real-time object detection. © (2024), (International Association of Engineers). All rights reserved.

关键词： Feature extraction

Multi-lesion Segmentation of Fundus Images using Improved UNet++

学校读者我要写书评

暂无评论

IAENG International Journal of computer science 2024年第10期51卷 1587-1595页

作者： Jiang, Haoyan Zhao, Ji School of Computer Science and Software Engineering University of Science and Technology Liaoning Anshan114051 China

Diabetic Retinopathy is a common microvascular complication of diabetes, and early and accurate diagnosis is crucial for minimizing its impact on vision. To address the complexity and diversity of lesions in diabetic retinopathy, as well as the presence of numerous small-scale lesions, this study proposes a multi-lesion segmentation framework based on an improved UNet++ architecture. Utilizing ResNet50 as the backbone network for feature extraction, we integrated a hybrid attention module into the residual block to enhance the model’s feature extraction capability in handling the complexity of lesions. To address the information loss of small lesions during feature extraction, we introduced and adapted Across Feature Map Attention as an auxiliary branch, which enhances the segmentation accuracy of small lesions. Furthermore, considering the insufficient feature extraction capability for DR lesions in shallow network layers, the model abandoned the deep supervision structure of traditional UNet++. Experiments employed a weighted hybrid loss function. Evaluations conducted on IDRiD and DDR segmentation datasets demonstrated effective segmentation of four typical Diabetic Retinopathy lesions. Results indicated that compared with other research methods, our approach achieved superior performance in Dice Coefficient and IoU metrics. © (2024), (International Association of Engineers). All rights reserved.

关键词： Semantic Segmentation