检索结果-内蒙古大学图书馆

Real-Time Employee Monitoring System using CCTV Cameras

Real-Time Employee Monitoring System using CCTV Cameras

2023 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems, ICSES 2023

作者： Harshavardhan, G. Singh, Nongmeikapam Thoiba Chanu, Asem Debala Chandigarh University Department of Computer Science and Engineering Punjab India Indian Institute of Information Technology Department of Computer Science and Engineering Assam Guwahati India

ISBN: (纸本)9798350319200

In this approach, we detail an approach to monitoring employees using preexisting setups such as CCTV camera feeds and a network connection. With this approach, we try to lower the system requirements to make it available to a wide range of setups. By relying on a software-based approach, employers can try it out and decide whether they wish to continue to subscribe, unlike a hardware-based approach, where they are locked onto the plan after the installation of the hardware. This approach will be modular in nature, where certain steps can be swapped out for more effective ones. It entails security personnel watching live CCTV footage. Instead of looking for anomalies, they might occasionally examine employees to ensure they are working, thus acting as remote supervisors. This function can be automated to create a more streamlined approach. This approach will predominantly use an employee's estimated position to predict whether they are working. This will depend on the type of job that they have. Through frequent sampling at regular intervals, we can estimate the amount of time an employee has spent working. This is a measure of the time spent working and not an employee's productivity. Though this model provides a basis for further derivative models, it has things that can be improved. One such factor would be to add further evaluation metrics to make the model more precise. © 2023 IEEE.

关键词： binary classifier CCTV Camera pose estimation real time employee monitoring technology (REMT)

来源：评论

学校读者我要写书评

暂无评论

Apple Leaf Disease Detection: Machine Learning & Deep Learning Techniques

Apple Leaf Disease Detection: Machine Learning & Deep Learni...

引用

2023 Intelligent Computing and Control for engineering and Business Systems, ICCEBS 2023

作者： Sebastian, Merin Suchithra, M.S. Antony, Chippy Maria Department of Computer Science Kerala Irinjalakuda India Indian Institute of Information Technology Department of Computer Science and Engineering Kerala Kottayam India

ISBN: (纸本)9798350394580

Early leaf disease detection is essential to mitigate potential crop loss and reduce reliance on chemical treatments. Here we have tried out several Machine Learning and Deep Learning Techniques for automatically detecting apple leaf diseases. Machine Learning techniques like Support Vector Machine, Random Forest, Naive Bayes, K-Nearest Neighbor, Decision Tree, and some Deep Learning methods like Sequential, and VGG-16. We have compared all these models and found that VGG-16 has more accurate results than the other models we used. The recommended model proposes profound figuring out how to improve the arrangement of apple leaf illness. Additionally, it demonstrates the highest validation accuracy of 97.23 percent on the dataset for apple leaf disease. Contrasted with some present futuristic, this approach performs better. Therefore, farmers can use the suggested system to streamline the process of classifying apple leaf disease and aid in the disease's early diagnosis and treatment. © 2023 IEEE.

关键词： apple leaf disease deep learning machine learning VGG-16

来源：评论

学校读者我要写书评

暂无评论

LiDAR 3D object detection based on improved PillarNet 9

LiDAR 3D object detection based on improved PillarNet

引用

9th International Conference on Intelligent Informatics and BioMedical sciences, ICIIBMS 2024

作者： Huang, Xinde Shanghai Institute of Technology School of Computer Science and Information Engineering Shanghai China

ISBN: (纸本)9798350363043

To address the limitations of current methods in detecting small objects, such as pedestrians and cyclists, within autonomous driving scenarios, we propose a novel 3D object detection algorithm based on an improved PillarNet architecture. These limitations often manifest as reduced recognition accuracy and increased susceptibility to false positives and negatives. The proposed method helps to capture key contextual information and spatial relationships between point cloud columns by adding a global attention mechanism to the pair. In addition, the addition of residual convolution helps to enhance the feature extraction capability of the network This comprehensive strategy expands the receptive field for point cloud features and bolsters the model's robustness in recognizing small objects within intricate environments. The effectiveness of the proposed method is validated through evaluations on the KITTI dataset. Compared to the baseline PillarNet, our improved algorithm achieves a 2.5% increase in the average detection accuracy for the cyclist category, demonstrating the efficacy of our approach. © 2024 IEEE.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

TSCompiler: efficient compilation framework for dynamic-shape models

引用

science China(information sciences) 2024年第10期67卷 67-84页

作者： Xiang LUO Chen ZHANG Chenbo GENG Yanzhi YI Jiahui HU Renwei ZHANG Zhen ZHANG Gianpietro CONSOLARO Fan YANG Tun LU Ning GU Li SHANG School of Computer Science Fudan University School of Electronic Information and Electrical Engineering Shanghai Jiao Tong University School of Computer Science and Technology Harbin Institute of Technology Huawei Technologies Co. Ltd. Huawei Paris Research Center School of Microelectronics Fudan University

Today's deep learning models face an increasing demand to handle dynamic shape tensors and computation whose shape information remains unknown at compile time and varies in a nearly infinite range at runtime. This shape dynamism brings tremendous challenges for existing compilation pipelines designed for static models which optimize tensor programs relying on exact shape values. This paper presents TSCompiler, an end-to-end compilation framework for dynamic shape models. TSCompiler first proposes a symbolic shape propagation algorithm to recover symbolic shape information at compile time to enable subsequent optimizations. TSCompiler then partitions the shape-annotated computation graph into multiple subgraphs and fine-tunes the backbone operators from the subgraph within a hardware-aligned search space to find a collection of high-performance schedules. TSCompiler can propagate the explored backbone schedule to other fusion groups within the same subgraph to generate a set of parameterized tensor programs for fused cases based on dependence analysis. At runtime, TSCompiler utilizes an occupancy-targeted cost model to select from pre-compiled tensor programs for varied tensor shapes. Extensive evaluations show that TSCompiler can achieve state-of-the-art speedups for dynamic shape models. For example, we can improve kernel efficiency by up to 3.97× on NVIDIA RTX3090, and 10.30× on NVIDIA A100 and achieve up to five orders of magnitude speedups on end-to-end latency.

关键词： machine learning tensor compilers dynamic shape operator fusion code generation auto-tuning

来源：评论

学校读者我要写书评

暂无评论

Despeckling SAR Images Using CNN-Based Approach Incorporating GAN and Gradient Estimation 27th

Despeckling SAR Images Using CNN-Based Approach Incorporatin...

引用

27th International Conference on Pattern Recognition, ICPR 2024

作者： Saha, Anirban Maji, Suman Kumar Department of Computer Science and Engineering Indian Institute of Technology Patna Bihar Patna801106 India

ISBN: (纸本)9783031783111

Synthetic Aperture Radar (SAR) technology stands at the forefront of capturing and processing Earth’s surface visuals due to its widespread acceptance across various organizations. However, the presence of unwanted random granular interference, commonly referred to as "speckle," poses a significant challenge in SAR data processing. Addressing this challenge, known as "despeckling," is crucial for extracting clear SAR visuals. This article introduces a novel CNN-based approach for despeckling SAR visuals contaminated with speckle. Our proposed model integrates a Generative Adversarial Network (GAN) module to estimate the distribution of contaminating speckle components from the input SAR data. Concurrently, a gradient estimator module captures the crisp changes in textural information within the input data. Subsequently, the input SAR data, the estimated noise distribution, and the extracted gradient undergo further processing through a deep convolutional module to generate a clean SAR visual. Unlike traditional methods that focus solely on learning the residual noisy component or the clean data, our proposed despeckling model learns the degradation pattern caused by noisy components while emphasizing gradient information, thereby capturing critical minute information. Experimental results demonstrate that our methodology significantly enhances despeckling performance compared to existing technologies in the literature. This research presents a promising step forward in advancing SAR visual despeckling techniques, with implications for improved data quality and interpretation in various applications. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Machine Learning-based Disease Diagnosis in Healthcare: A Focus on Brain Tumor Detection

Machine Learning-based Disease Diagnosis in Healthcare: A Fo...

引用

2023 International Conference on Recent Advances in science and engineering technology, ICRASET 2023

作者： Singh, Nongmeikapam Thoiba Kaur, Charnpreet Kaur, Puneet Kumari, Navjot Chanu, Asem Debala Chandigarh University Department of Computer Science and Engineering Punjab India Indian Institute of Information Technology Department of Computer Science and Engineering Assam Guwahati India

ISBN: (纸本)9798350306927

One of the most crucial concerns facing society is healthcare for individuals. It searches out the most accurate and trustworthy illness diagnosis possible to make sure that patients get the care they need as soon as possible. Because this identification is usually sophisticated, other fields like statistics and computer science are needed for the search-related health component. To adopt cutting-edge approaches, these disciplines must go beyond the conventional ones. The actual number of new strategies makes it feasible to provide a thorough summary without getting into specifics. We suggest a detailed investigation of illnesses connected to machine learning in humans to achieve this. This study focuses on existing machine learning growth techniques used in the medical profession to diagnose human ailments in order to find fascinating trends, make unimportant forecasts, and aid in decision-making. Preparing the data is the initial step in every machine learning problem. This procedure involves the use of brain tumor classification datasets for training and testing a model. The dataset comprises four types of human brain tumor images, namely glioma tumor (gt), meningioma tumor (mt), no tumor (nt), and pituitary tumor (pt). The primary stage of a machine learning project is data cleaning. The quality of our data determines how well our machine learning model performs. Before supplying the data to the model for training, cleaning is always necessary. When the information has been gathered and sanitised. These cleaned data will be used to train the Support Vector Classifier (SVM), Linear Regression Classifier, and Logistic Regression Classifier. Using a confusion matrix, the models' quality will be evaluated. After training all the models, the predictions from each model will be combined to anticipate the disease based on the symptoms provided as input. In this proposed method, no tumor result achieved almost 100 percent accuracy in this specific dataset. Pituitary tum

关键词： Tumors

来源：评论

学校读者我要写书评

暂无评论

Parametrized Optimization Based on an Investigation of Musical Similarities Using SPARK and Hadoop

引用

SN computer science 2024年第1期5卷 78页

作者： Chaudhary, Savita Karthik, V. Shankar, R. Taranum, Ayesha Naresh, E. Department of Computer Science and Engineering Sir M Visvesvaraya Institute of Technology Bengaluru Bengaluru India Department of Information Science and Engineering Ramaiah Institute of Tech B’lore India Department of Computer Science and Engineering BMS Institute of Technology and Engineering Bengaluru India Department of Computer Science and Engineering Vidyavardhaka College of Engineering Mysuru India Department of Information Technology Manipal Institute of Technology Bengaluru Manipal Academy of Higher Education Manipal India

The big data processing framework Spark is used to power a parameterizable recommender system that can make recommendations for music based on a user’s individual tastes and take into account a variety of musical tonal qualities. The system as it is presently built is completely scalable, which means that additional songs can be contributed to the data, the cluster size could be increased, and new types of audio information, in addition to more cutting-edge similarity evaluations, may be included. Another issue discussed in this research paper is the parallel collection of required audio characteristics on a computer cluster. Song recommendations for a dataset including more than 114,000 songs may be created on a Spark cluster with 16 nodes in under 12 s by integrating eight distinct audio feature types and similarity assessments. After the features have been retrieved, they are sent to the Spark-based recommender system to be processed. The calculated distance was displayed, examined, and graphically depicted. By computing the distance depending on the melody, rhythmic, and timbral components of the music, the final software controls song suggestion. © 2023, The Author(s).

关键词： Big data Hadoop Music information retrieval Optimization Spark

来源：评论

学校读者我要写书评

暂无评论

Pedestrian Multi-Object Tracking Algorithm Based on Fish Eye Camera Distortion Correction 9

Pedestrian Multi-Object Tracking Algorithm Based on Fish Eye...

引用

9th International Conference on Intelligent Informatics and BioMedical sciences, ICIIBMS 2024

作者： Xia, Tian Shanghai Institute of Technology School of computer science and Information Engineering Shanghai China

ISBN: (纸本)9798350363043

This paper proposes a multi-object tracking algorithm for distorted video images captured by a fisheye camera. The method addresses the issue of residual shortening distortion in calibrated pedestrian images. It utilizes image data augmentation techniques to create a pedestrian image dataset with calibrated distortion features, thereby improving the network's recognition ability for such erroneous images. Experimental results demonstrate that the network trained using this dataset exhibits a 1.37% improvement in accuracy, a 1.42% reduction in missed detection rate, and a 0.95% reduction in false detection rate compared to other algorithms in recognizing fisheye camera distortion features. Additionally, based on YoloV5, the decoupled head structure is redesigned using the Hybrid Channels strategy, and the CSPStackRep Block is used as the backbone network. Combining the laboratory's domain-controlled hardware conditions, a hardware-friendly strategy is employed to optimize and adjust the network structure, facilitating hardware deployment and achieving higher efficiency and lower latency. Experimental results show that compared to other algorithms, this approach can improve MOTA by 1.47% and MOTP by 0.51% under a fisheye camera while maintaining the same frame rate, thereby enhancing the tracking performance. © 2024 IEEE.

关键词： Image enhancement

来源：评论

学校读者我要写书评

暂无评论

Weakly Supervised Salient Object Detection with Dual-path Mutual Reinforcement Network 9

Weakly Supervised Salient Object Detection with Dual-path Mu...

引用

9th International Conference on Intelligent Informatics and BioMedical sciences, ICIIBMS 2024

作者： Ou, Yanling Shi, Yanjiao Shanghai Institute of Technology School of Computer Science and Information Engineering Shanghai China

ISBN: (纸本)9798350363043

Traditional salient object detection (SOD) methods heavily rely on large-scale pixel-level datasets, making them both time-consuming and expensive. However, it is a significant challenge to effectively integrate long-range dependencies and local feature information for weakly supervised salient object detection (WSOD). In this paper, we propose a novel dual-path mutual reinforcement network (DMRN) for WSOD, which includes a CNN branch and a ViT branch in the encoder to effectively capture local details and global features, respectively. A mutual fusion module (MFM) is introduced to perform efficient feature fusion and establish long-range contextual dependencies. Additionally, an attention-based cross-feature module (ACFM) is designed to enhance feature interactions between branches, ensuring accurate localization of salient objects. To further enhance the model, two auxiliary prediction modules (APM) are used to generate pseudo labels from different perspectives, compensating for sparse annotations. Experimental results on five benchmark datasets show that the proposed method outperforms 11 existing weakly supervised methods and shows its superior performance in capturing salient features and achieving accurate object detection. © 2024 IEEE.

关键词： Object recognition

来源：评论

学校读者我要写书评

暂无评论

AcKnowledge: Acquired Knowledge Representation by Small Language Model Without Pre-training 1

AcKnowledge: Acquired Knowledge Representation by Small Lang...

引用

1st Workshop on Towards Knowledgeable Language Models, KnowLLM 2024 - Co-located with ACL 2024

作者： Das, Sourav Chatterji, Sanjay Mukherjee, Imon Department of Computer Science and Engineering Indian Institution of Information Technology Kalyani West Bengal Kalyani India

ISBN: (纸本)9798891761544

Large language models (LLMs) are pre-trained on enormous amounts of text data and show acclaimed success in knowledge representation. However, there are two bottlenecks with this approach. (1) Pre-training data cannot be regularly updated once the models are deployed, and it is not very fruitful if the model cannot represent updated knowledge. (2) The consistently increasing size and computational resources make it difficult for noncommercial and individual researchers to fine-tune and scale these language models. Major LLMs with external knowledge are also proprietary. In this paper, we propose AcKnowledge, a framework wrapped around a small, non-pre-trained language model for an open-domain question-answering (QA) experiment. AcKnowledge learns relevant knowledge from the internet via meta-learning based on user questions, and re-learns from user feedback if knowledge is misrepresented. Our efficient knowledge representation framework avoids pre-training overhead while enabling updated information. Benchmarking shows competitive performance against similarly sized state-of-the-art (SoTA) LLMs on gold standard QA datasets, demonstrating the potential of integrating internet search and user feedback for improved performance and general-izability. The repository of the work is available at https://***/SouravD-Me/AcKnowledge-KnowledgeLM-ACL-2024. © 2024 Association for Computational Linguistics.

关键词： Knowledge representation

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：