检索结果-内蒙古大学图书馆

7th International Conference on Automation Electronics and Electrical Engineering

作者： Qian Ping Zhang Yong Cao Yang Zhou Gang Qi Zhongyi State Grid Dongtou Power Supply Co Dept Management Hangzhou Peoples R China State Grid Zhejiang Elect Power Dept Equipment Hangzhou Peoples R China State Grid Jiaxing Power Supply Co Dept Operat & Maintenance Jiaxing Zhejiang Peoples R China State Grid Zhejiang Elect Power Jiaxing Power Sup Dept Operat & Inspect Jiaxing Zhejiang Peoples R China State Grid Zhejiang Elect Power Jiaxing Power Sup Dept Operat Jiaxing Zhejiang Peoples R China

ISBN: (数字)9798350377033

ISBN: (纸本)9798350377040;9798350377033

Electric robot will obtain a large amount of image information during inspection, and if these images are checked whether there are faults in power inspection is time-consuming and labor-intensive. There is an urgent need for power image Chinese title generation technology to solve it. However, existing image Chinese title generation methods face the problems of small training data sets, differences in specific applications, and few methods for generating Chinese titles for power images. To this end, this paper proposes a self-supervised learning-based image Chinese title generation algorithm for fault detection in electric robot inspection. Specifically, a contrastive learning-based model to automatically capture the semantic relationship between images and text. Then, we propose an end-to-end encoding-decoding model combined with an attention mechanism to obtain Chinese title generation for inspection images. The effectiveness of the proposed algorithm is experimentally verified on two real datasets.

关键词： self-supervised learning attention mechanism encoder-decoder image caption generation electric robot Semantic feature

来源：评论

学校读者我要写书评

暂无评论

GAP: A Grammar and Position-Aware Framework for Efficient Recognition of Multi-Line Mathematical Formulas 24

GAP: A Grammar and Position-Aware Framework for Efficient Re...

引用

17th ACM International Conference on Web Search and Data Mining (WSDM)

作者： Yang, Zhe Liu, Qi Zhang, Kai Tong, Shwei Chen, Enhong Univ Sci & Technol China Sch Comp Sci & Technol Hefei Anhui Peoples R China State Key Lab Cognit Intelligence Hefei Anhui Peoples R China

ISBN: (纸本)9798400703713

Formula recognition endeavors to automatically identify mathematical formulas from images. Currently, the encoder-decoder model has significantly advanced the translation from image to corresponding formula markups. Nonetheless, previous research primarily concentrated on single-line formula recognition, ignoring the recognition of multi-line formulas, which presents additional challenges such as more stringent grammatical restrictions and twodimensional positions. In this work, we present GAP (Grammar And Position-Aware formula recognition), a comprehensive framework designed to tackle the challenges in multi-line mathematical formula recognition. First, to overcome the limitations imposed by grammar, we design a novel Grammar Aware Contrastive Learning (GACL) module, integrating complex grammar rules into the transcription model through a contrastive learning mechanism. Furthermore, primitive contrastive learning lacks clear directions for comprehending grammar rules and can lead to unstable convergence or prolonged training cycles. To enhance training efficiency, we propose Rank-Based Sampling (RBS) specialized for multi-line formulas, which guides the learning process by the importance ranking of different grammar errors. Finally, spatial location information is critical considering the two-dimensional nature of multiline formulas. To aid the model in keeping track of that global information, we introduced a Visual Coverage (VC) mechanism that incorporates historical attention information into the image features via a parameter-free way. To validate the effectiveness of our GAP framework, we construct a new dataset Multi-Line containing 12,002 multi-line formulas and conduct extensive experiments to show the efficacy of our GAP framework in capturing grammatical rules, enhancing recognition accuracy, and enhancing training efficiency. Codes and datasets are available at https://***/Sinon02/GAP.

关键词： Formula Recognition Contrastive Learning encoder-decoder

来源：评论

学校读者我要写书评

暂无评论

Methodology for Video Description Generation with Semantic Feature Selection and Attention Fusion 5

Methodology for Video Description Generation with Semantic F...

引用

5th International Conference on Computer Engineering and Application (ICCEA)

作者： Che, Zhanbin Wang, Zhihao Zhongyuan Univ Technol Sch Comp Sci Zhengzhou Peoples R China

ISBN: (纸本)9798350386783;9798350386776

To enhance the semantic relevance between the textual descriptions of entities and the video content and to address the inefficiency of multi-modal features. This paper proposes a novel method for video description generation focusing on semantic feature selection and attention fusion. By selecting semantic objects from the video that appear in the text description and retrieving the corresponding behavioral representations from the motion features. Utilizing a cross-attention mechanism, integrating semantic and action features. The fused features are then input into a decoder to generate semantic descriptions. Experiments demonstrate that this approach effectively screens semantic objects within videos and generates complete textual descriptions based on their actions.

关键词： Video description Semantic feature selection Attention fusion encoder-decoder

来源：评论

学校读者我要写书评

暂无评论

Multimodal Enhanced Target Representation for Machine Translation 1

引用

13th International Conference on Computer Engineering and Networks (CENet)

作者： Zang, Xiaogang Zhu, Huidong Dai, Xue Zhengzhou Univ Light Ind Zhengzhou Peoples R China

ISBN: (数字)9789819992430

ISBN: (纸本)9789819992454;9789819992430;9789819992423

Recently, visual annotation is successfully introduced into the encoder-decoder framework to open a new multimodal neural machine translation (MMT) scenario. However, it is still difficult for MMT to simulate the ground-truth target future context due to the autoregressive decoder. The visual annotation, which presents the content described in the bilingual parallel sentence pair, provides a ground-truth global contextual information for the auto-regressive decoder. Thus, we propose to directly learn a practical target for future contextual information from the visual modality space, thereby enhancing the target representation of the decoder. Empirical results on several widely used multimodal translation benchmarks demonstrated that the proposed approach significantly improved the performance of MMT over strong baselines.

关键词： Neural machine translation encoder-decoder Multimodal

来源：评论

学校读者我要写书评

暂无评论

An Attention Model Based Heuristic Algorithm for Storage Selection and Order Picking Path Joint Optimization 27

An Attention Model Based Heuristic Algorithm for Storage Sel...

引用

27th International Conference on Computer Supported Cooperative Work in Design (CSCWD)

作者： Chen, Feng Wang, Li Zhang, Jiawei Sun, Qi Lai, Pinyan State Grid Zhejiang Elect Power CO LTD Mat Branch Quzhou Zhejiang Peoples R China Shanghai Jiao Tong Univ Shanghai Peoples R China

ISBN: (纸本)9798350349184;9798350349191

This research introduces an advanced attention-driven model designed to optimize warehouse order-picking, an enhancement of the foundational work by Kool et al. (2018). Fundamentally, our model incorporates an enhanced masking mechanism and context-aware decoder, streamlining the order-picking process. In essence, our model presents an attention model based heuristic solution to the long-standing problem of order-picking optimization, leveraging the latest in attention-based deep learning and reinforcement learning techniques. Compared to existing methods, our innovative model shows superior performance, offering significant potential in warehousing solutions.

关键词： order picking attention model storage selection reinforcement learning encoder-decoder

来源：评论

学校读者我要写书评

暂无评论

Hyperparameter optimization for deep neural network models: a comprehensive study on methods and techniques

引用

INNOVATIONS IN SYSTEMS AND SOFTWARE ENGINEERING 2023年 1-12页

作者： Roy, Sunita Mehera, Ranjan Pal, Rajat Kumar Bandyopadhyay, Samir Kumar Univ Calcutta Dept Comp Sci & Engn JD-2Sect 3Salt Lake Kolkata 700106 West Bengal India Anodot Inc 20130 Lakeview Ctr PlSuite 400 Ashburn VA 20147 USA Bhawanipur Educ Soc Coll 5Elgin RdSreepallyBhowanipore Kolkata 700020 West Bengal India

Advancements in computing and storage technologies have significantly contributed to the adoption of deep learning (DL)-based models among machine learning experts. Although a generic model can be used in the search for a near-optimal solution in any problem domain, what makes these DL models context-sensitive is the combination of the training data and the hyperparameters. Due to the lack of inherent explainability of DL models, the hyperparameter optimization (HPO) or tuning specific to each model is a combination of art, science, and experience. In this article, we have explored various existing methods or ways to identify the optimal set of values for the hyperparameters specific to the DL models along with the techniques to realize those methods in real-life situations. The article also includes a detailed comparative study among various state-of-the-art HPO techniques using the Keras Tuner tuning toolkit and highlights the observations describing how the model performance can be improved by applying various HPO techniques.

关键词： Automated machine learning Brain tumour segmentation Deep learning encoder-decoder Hyperparameter optimization SA-Net

来源：评论

学校读者我要写书评

暂无评论

GRAN: graph recurrent attention network for pedestrian orientation classification

引用

COMPLEX & INTELLIGENT SYSTEMS 2023年第1期9卷 891-908页

作者： Li, Xiao Ma, Shexiang Shan, Liqing Liu, Sheng Chai, Song Li, Xiao Tianjin Univ Technol Sch Comp Sci & Engn Tianjin 300384 Peoples R China Tianjin Univ Technol Sch Integrated Circuit Sci & Engn Tianjin 300384 Peoples R China Southeast Univ Sch Informat Sci & Engn Nanjing 211189 Peoples R China

In complex traffic scenes, accurate identification of pedestrian orientations can help drivers determine pedestrian trajectories and help reduce traffic accidents. However, there are still many challenges in pedestrian orientation recognition. First, due to the irregular appearance of pedestrians, it is difficult for general Convolutional Neural Networks (CNNs) to extract discriminative features. In addition, more features of body parts help to judge the orientation of pedestrians. For example, head, arms and legs. However, they are usually small and not conducive to feature extraction. Therefore, in this work, we use several discrete values to define the orientation of pedestrians, and propose a Gated Graph Neural Network (GGNN)-based Graph Recurrent Attention Network (GRAN) to classify the orientation of pedestrians. The contributions are as follows: (1) We construct a body parts graph consisting of head, arms and legs on the feature maps output by the CNN backbone. (2) Mining the dependencies between body parts on the graph via the proposed GRAN, and utilizing the encoder-decoder to propagate features among graph nodes. (3) In this process, we propose an adjacency matrix with attention edge weights to dynamically represent graph node relationships, and the edge weights are learned during network training. To evaluate the proposed method, we conduct experiments on three different benchmarks (PDC, PDRD, and Cityscapes) with 8, 3, and 4 orientations, respectively. Note that the orientation labels for PDRD and Cityscapes are annotated by our hand. The proposed method achieves 97%, 91% and 90% classification accuracy on the three data sets, respectively. The results are all higher than current state-of-the-art methods, which demonstrate the effectiveness of the proposed method.

关键词： Pedestrian orientation Body parts Graph Attention encoder-decoder

来源：评论

学校读者我要写书评

暂无评论

The Impact of Text Generation Techniques on Neural Image Captioning: An Empirical Study

引用

INTERNATIONAL JOURNAL OF ADVANCED COMPUTER SCIENCE AND APPLICATIONS 2023年第10期14卷 1080-1090页

作者： Ding, Linna Jiang, Mingyue Nie, Liming Qing, Zuzhang Ding, Zuohua Zhejiang Sci Tech Univ Fac Comp Sci & Technol Hangzhou Zhejiang Peoples R China Nanyang Technol Univ Sch Comp Sci & Technol Singapore Singapore Zhejiang Petr Integrated Energy Sales Co Ltd Hangzhou Zhejiang Peoples R China

Image captioning is an advanced NLP task that has various practical applications. To meet the requirement of visual information understanding and textual information generation, the encoder-decoder framework has been widely adopted by image captioning models. In this context, the encoder is responsible for transforming an image into vector representation, and the decoder acts as a text generator for yielding an image caption. It is obvious and intuitive that the decoder is crucial for the entire image captioning model. However, there is a lack of comprehensive studies in which the impact of various aspects of the decoder on the image captioning is investigated. To advance the understanding of the impacts of text generation techniques employed by the decoder, we conduct an extensive empirical analysis of three types of language models, two types of decoding strategies and two types of training methods, based on four stateof-the-art image captioning models. Our experimental results demonstrate that the language model affects the performance of image captioning models, while different language models may benefit different image captioning models. In addition, it is also revealed that among the decoding and training strategies under investigation, the beam search, AOA mechanism and the reinforcement learning based training method can generally improve the performance of image captioning models. Moreover, the results also show that the combinational usage of these strategies always outperforms the use of single strategy for the task of image captioning.

关键词： Image captioning encoder-decoder text generation techniques

来源：评论

学校读者我要写书评

暂无评论

Satellite imagery analysis for road segmentation using U-Net architecture

引用

JOURNAL OF SUPERCOMPUTING 2022年第10期78卷 12710-12725页

作者： Chaudhary, Vidhi Buttar, Preetpal Kaur Sachan, Manoj Kumar St Longowal Inst Engn & Technol Dept Comp Sci & Engn Sangrur Punjab India

Road network plays a significant role in today's urban development. These are a vital part of applications such as automatic road navigation, traffic management, route optimization, etc. In this paper, we aim to explore the potential and performance of convolution neural network architecture, U-Net, performing semantic segmentation to detect roads. The model has been evaluated on Massachusetts Roads Dataset with the estimation of numerous parameters such as filter stride, learning rate, training epochs, data size, and various augmentation techniques. The employment of the proposed technique and hyperparameters to the U-Net architecture gives a boost to test accuracy and improves the dice coefficient by a value of 0.007. Experimental results thereby demonstrate that our model is computationally efficient and achieves comparable segmentation results.

关键词： Convolutional neural network (CNN) Road extraction Satellite images Fully convolutional networks Semantic image segmentation encoder-decoder

来源：评论

学校读者我要写书评

暂无评论

Pre-Training-Based Grammatical Error Correction Model for the Written Language of Chinese Hearing Impaired Students

引用

IEEE ACCESS 2022年 10卷 35061-35072页

作者： Chen, Binbin Zhang, Jingyu Qiannan Normal Univ Nationalities Sch Int Studies Duyun 558000 Guizhou Peoples R China Xian Int Studies Univ English Literature Inst Xian 710128 Shaanxi Peoples R China

Grammatical error correction has been considered as an application closely related to daily life and an important shared task in many prestigious competitions and workshops. The neural machine translation with an encoder-decoder architecture containing language models has been the fundamental solution for the grammatical error correction. Whereas Grammatical error correction task on texts of hearing impaired people or its solution has not been seen yet, and common Grammatical error correction tasks are suffering several challenges, such as insufficient training data, insufficient accuracy due to the unsatisfactory capacity of extracting semantic and grammatical patterns. Under these circumstances, we proposed a novel encoder-decoder architecture based on multi-head self-attention along with multiple strategies, which excels at extracting deep representations from the corrupted sentences of hearing impaired students and further reconstructing the sentences into grammatical ones. Via the re-ranking strategy, our model can correct various kinds of errors including spelling and complex syntax errors. The ablation experiments prove that the semantic extracting of self-attention mechanism excluding the position encoding with the word order shuffle operation can significantly learn the hearing impaired students' sentence patterns whose word order is quite different from the ones of hearing people and improve the correction scores. The pre-training can enhance the restoring efficiency of sentence structure in the decoding process. The comparison experiments with baseline models show that our model obtains superior performance either in the hearing impaired students' grammatical error correction or in a common grammatical error correction shared task.

关键词： Task analysis Error correction Decoding Machine translation Training data Training Transformers Hearing impaired student grammatical error correction self-attention encoder-decoder pre-training

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：