检索结果-内蒙古大学图书馆

ieee Winter Applications and computer vision workshops (WACVW)

作者： Lukas Hüttner Martin Mayr Thomas Gorges Fei Wu Mathias Seuret Andreas Maier Vincent Christlein Pattern Recognition Lab Friedrich-Alexander Universität Erlangen-Nürnberg Erlangen Germany

ISBN: (数字)9798331536626

ISBN: (纸本)9798331536633

The continuous expansion of neural network sizes is a notable trend in machine learning, with transformer models exceeding 20 billion parameters in computer vision. This growth comes with rising demands for computational resources and large-scale datasets. Efficient techniques for transfer learning thus become an attractive option in setups with limited data, as in handwriting recognition. Recently, parameter-efficient fine-tuning (PEFT) methods, such as low-rank adaptation (LoRA) and weight-decomposed low-rank adaptation (DoRA), have gained wide-spread interest. In this paper, we explore tradeoffs in parameter-efficient transfer learning using the synthetically pretrained Transformer-Based Optical Character recognition (TrOCR) model for handwritten text recognition with LoRA and DoRA. Additionally, we analyze the performance of full fine-tuning with a limited number of samples, scaling from a few-shot learning scenario up to using the whole dataset. We conduct experiments on the popular IAM Handwriting database as well as the historical READ 2016 dataset. We find that (a) LoRA/DoRA does not outperform full fine-tuning as opposed to a recent paper and (b) LoRA/DoRA is not substantially faster than full fine-tuning of TrOCR.

关键词： Training Handwriting recognition computer vision Adaptation models Runtime Text recognition Databases Transfer learning Training data Transformers

来源：评论

学校读者我要写书评

暂无评论

Gate-Shift-Pose: Enhancing Action recognition in Sports with Skeleton Information

Gate-Shift-Pose: Enhancing Action Recognition in Sports with...

引用

ieee Winter Applications and computer vision workshops (WACVW)

作者： Edoardo Bianchi Oswald Lanz Free University of Bozen-Bolzano Piazza Università 1 Bolzano Italy

ISBN: (数字)9798331536626

ISBN: (纸本)9798331536633

This paper introduces Gate-Shift-Pose, an enhanced version of Gate-Shift-Fuse networks, designed for athlete fall classification in figure skating by integrating skeleton pose data alongside RGB frames. We evaluate two fusion strategies: early-fusion, which combines RGB frames with Gaussian heatmaps of pose keypoints at the input stage, and latefusion, which employs a multi-stream architecture with attention mechanisms to combine RGB and pose features. Experiments on the FR-FS dataset demonstrate that Gate-Shift-Pose significantly outperforms the RGB-only baseline, improving accuracy by up to 40% with ResNet18 and 20% with ResNet50. Early-fusion achieves the highest accuracy (98.08%) with ResNet50, leveraging the model's capacity for effective multimodal integration, while latefusion is better suited for lighter backbones like ResNet18. These results highlight the potential of multimodal architectures for sports action recognition and the critical role of skeleton pose information in capturing complex motion patterns.

关键词： Heating systems Accuracy conferences computer architecture Logic gates Skeleton Multisensory integration pattern recognition Sports Residual neural networks

来源：评论

学校读者我要写书评

暂无评论

Guest Editors' Introduction to the Special Section on Award Winning Papers from the ieee CS conference on computer vision and pattern recognition (CVPR)

引用

ieee TRANSACTIONS ON pattern ANALYSIS AND MACHINE INTELLIGENCE 2009年第12期31卷 2113-2114页

作者： Boyer, Kim Shah, Mubarak Syeda-Mahmood, Tanveer Rensselaer Polytech Inst Dept Elect Comp & Syst Engn Troy NY 12180 USA Univ Cent Florida Sch Elect Engn & Comp Sci Orlando FL 32816 USA IBM Almaden Res Ctr San Jose CA 95120 USA

The three articles in this special section are selected papers from the ieee CS conference on computer vision and pattern recognition that was held in Anchorage, AL, in June 2008.

关键词： computer vision pattern recognition Awards Stereo Image Processing Stereo vision Image Reconstruction computer Society History Surface Reconstruction Layout

来源：评论

学校读者我要写书评

暂无评论

Special Editors' Introduction to the Special Issue on Award-Winning Papers from the ieee conference on computer vision and pattern recognition 2010 (CVPR 2010)

引用

ieee TRANSACTIONS ON pattern ANALYSIS AND MACHINE INTELLIGENCE 2012年第9期34卷 1665-1666页

作者： Darrell, Trevor Hogg, David Jacobs, David Univ Calif Berkeley Div Comp Sci Berkeley CA 94704 USA Int Comp Sci Inst Berkeley CA 94704 USA Univ Leeds Sch Comp Leeds W Yorkshire England Univ Maryland Dept Comp Sci College Pk MD 20742 USA

The nine award-winning papers in this special section were presented at the ieee conference on computer vision and pattern recognition (CVPR 2010)that was held 13-18 June 2010 in San Francisco, CA.

关键词： Special Issues And Sections Meetings computer vision pattern recognition Awards

来源：评论

学校读者我要写书评

暂无评论

Guest Editors' Introduction to the Special Section on Award-Winning Papers from the ieee conference on computer vision and pattern recognition 2009 (CVPR 2009)

引用

ieee TRANSACTIONS ON pattern ANALYSIS AND MACHINE INTELLIGENCE 2011年第12期33卷 2339-2340页

作者： Essa, Irfan Kang, Sing Bing Pollefeys, Marc Georgia Inst Technol Sch Interact Comp Atlanta GA 30332 USA Microsoft Corp Redmond WA 98052 USA ETH Dept Comp Sci CH-8092 Zurich Switzerland

来源：评论

学校读者我要写书评

暂无评论

On the Automated Selection of Neural Network Architectures in the Task of pattern recognition for Intelligent Transportation Systems

On the Automated Selection of Neural Network Architectures i...

引用

2025 Systems of Signals Generating and Processing in the Field of on Board Communications, SOSG 2025

作者： Moseva, Alesya Gorodnichev, Mikhail Moseva, Marina Kharrasov, Kamil Moscow Technical University of Communications and Informatics Moscow Russia

ISBN: (纸本)9798331532635

Exponential growth in the number of computing resources, their cheapening, and the wide availability of large data sets have ensured a great demand for information systems based on neural networks. Such systems are widely used in many spheres of life: in medicine, detecting diseases at early stages and helping to design new drugs;in the financial sector, detecting fraudulent transactions and calculating risks, including the development of intelligent transportation systems. Intelligent transportation systems (ITS) actively use computer vision technologies to solve such tasks as object recognition, traffic analysis, traffic forecasting and road safety. However, the development of effective neural network models for these tasks requires significant resources and high qualification of specialists. The system of automated training and selection of neural network architecture can significantly simplify and speed up this process. Data collection, selection of the necessary model and tuning of hyperparameters requires specialized knowledge and skills. This work is aimed at developing a tool for automated selection of neural network architecture, which will accelerate the development and implementation of models, reduce development costs, improve the quality and reliability of solutions, provide scalability and adaptability to changing conditions. © 2025 ieee.

关键词： Intelligent systems

来源：评论

学校读者我要写书评

暂无评论

Classification of Colorectal Cancer Tissues Using Stacking Ensemble Learning 7th

Classification of Colorectal Cancer Tissues Using Stacking E...

引用

7th International conference on Communication, Device and Networking, ICCDN 2024

作者： Das, Abhrodeep Hazra, Animesh Department of Computer Science and Engineering Jalpaiguri Government Engineering College West Bengal Jalpaiguri735102 India

ISBN: (纸本)9789819764648

Advancement in digital pathology has enabled deep learning-based computer vision techniques for automated diagnosis and prognosis of diseases. The essentiality of early detection and prognosis of any cancer category lies in the fact that it can speed up the subsequent medical treatment procedures of patients. About 10% of all cancer cases worldwide are related to colorectal cancer (CRC), and it is also the third most common category of cancer (Egeblad et al. in Dev Cell 18:884–901, 2010). So, it is clinically important to classify and make an objective evaluation of colorectal cancer histological images. The classification performance of current methodologies primarily relies on the use of various combinations of texture-based features and classifiers or transfer learning to classify different organisational kinds. As a result of the diversity of tissue types and characteristics present in histological images, classification is still a challenging task. In this study, we have put forth a novel and effective stacking (Wolpert in Neural Netw 5:241–59, 1992) ensemble (Zhang and Yunqian (eds) Ensemble machine learning: methods and applications. Springer Science;Business Media, 2012) technique for classification on the histopathological image analysis benchmark dataset Kather-5 K (Kather et al. in Sci Rep 6:27988, 2016). The ensemble consists of two cutting-edge deep learning architectures, ResNet18 (He et al. in Proceedings of the ieee conference on computer vision and pattern recognition 2016, pp 770–778) and EfficientNetB0 (Tan and Le Efficientnet: rethinking model scaling for convolutional neural networks. International conference on machine learning 2019 May 24. PMLR, pp 6105–6114), acting as weak learners, and an ANN acting as the meta-learner. The proposed approach obtained a remarkable accuracy of 97.20% in CRC classification on the said dataset. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Colorectal cancer Convolutional neural network Ensemble learning Residual network Stacking

来源：评论

学校读者我要写书评

暂无评论

Offline Signature Verification in the Banking Domain

Offline Signature Verification in the Banking Domain

引用

ieee Winter Applications and computer vision workshops (WACVW)

作者： Valentina Arrigoni Group AI Products UniCredit Milan Italy

ISBN: (数字)9798331536626

ISBN: (纸本)9798331536633

In this paper, we address a specific industrial application of offline signature verification, in a context arising in the banking domain. In this context, we need to verify signatures placed in different types of real-world documents, e.g. bank cheques. Due to the huge amount of clients the problem is framed into the writer-independent configuration. Peculiar to this industrial application is that the signatures are associated with additional information, i.e. the personal data associated to the bank account. We propose a deep learning architecture to solve the signature verification of our industrial application, in which we take advantage of the additional textual information with a pretraining of an image embedding component driven by a text-recognition task. Moreover, our model employs an attention mechanism over the image embeddings to learn more powerful features. We extensively compare our model with a state-of-the-art siamese approach, with a text recognition model adopted to specifically solve this task and using a general purpose image encoder pretrained on the ImageNet dataset instead of text recognition embeddings. Our model proves to be more effective to solve the offline signature verification in two real-world large scale datasets. Ablation studies also confirm the importance of our ideas: the pretraining of the image encoders with a text recognition task and the attention mechanism.

关键词： Handwriting recognition Data privacy computer vision Attention mechanisms Text recognition conferences Deep architecture Banking computer architecture

来源：评论

学校读者我要写书评

暂无评论

Lights, Camera, Matching: The Role of Image Illumination in Fair Face recognition

Lights, Camera, Matching: The Role of Image Illumination in ...

引用

ieee Winter Applications and computer vision workshops (WACVW)

作者： Gabriella Pangelinan Grace Bezold Haiyu Wu Michael C. King Kevin W. Bowyer Florida Institute of Technology Melbourne FL USA University of Notre Dame Notre Dame IN

ISBN: (数字)9798331536626

ISBN: (纸本)9798331536633

Facial brightness is a key image quality factor impacting face recognition accuracy differentials across demographic groups. In this work, we aim to decrease the accuracy gap between the similarity score distributions for Caucasian and African American female mated image pairs, as measured by d’ between distributions. To balance brightness across demographic groups, we conduct three experiments, interpreting brightness in the face skin region either as median pixel value or as the distribution of pixel values. Balancing based on median brightness alone yields up to a 46.8% decrease in d’, while balancing based on brightness distribution yields up to a 57.6% decrease. In all three cases, the similarity scores of the individual distributions improve, with mean scores maximally improving 5.9% for Caucasian females and 3.7% for African American females.

关键词： Image quality computer vision Accuracy Face recognition conferences Brightness Lighting Licenses Cameras Skin

来源：评论

学校读者我要写书评

暂无评论

Multi-Modal Large Language Model Driven Augmented Reality Situated Visualization: The Case of Wine recognition

Multi-Modal Large Language Model Driven Augmented Reality Si...

引用

ieee Winter Applications and computer vision workshops (WACVW)

作者： Vincenzo Armandi Andrea Loretti Lorenzo Stacchio Pasquale Cascarano Gustavo Marfia Department of the Arts University of Bologna Bologna Italy Department of Political Sciences Communication and International Relations University of Macerata Macerata Italy

ISBN: (数字)9798331536626

ISBN: (纸本)9798331536633

Situated Visualizations (SV) and reality-based information retrieval systems, enhanced by Mixed Reality (MR) and Augmented Reality (AR), enable the overlay of digital information onto real-world objects, providing context-aware content through computer vision. Despite their potential, these systems face significant challenges in scalability and adaptability, particularly for domains like wine recognition, where diverse label designs, frequent updates, and limited historical databases complicate automated analysis. SOLLAMA (SOmmeLier LlAMA) is a novel wine recognition framework designed to address the scalability and adaptability challenges of AR systems in recognizing diverse wine labels. Leveraging Multimodal Large Language Models (MLLMs), SOLLAMA integrates visual and textual cues for accurate label interpretation, bypassing the need for extensive image datasets and traditional OCR methods. Built on the Augmented Wine recognition (AWR) system, it replaces the OCR module with LLAMA 3.2 for advanced text recognition and contextual understanding. Key benefits include scalability across diverse designs and simplified, server-free deployment. Experimental validation on a dataset of wine labels from Italy's Emilia-Romagna region highlights the system's effectiveness and its potential to transform wine recognition in AR-based applications.

关键词： Visualization computer vision Accuracy Scalability Large language models Face recognition Optical character recognition Information retrieval Robustness Augmented reality

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：