检索结果-南通市图书馆

Low rank optimization for efficient deep learning:making a balance between compact architecture and fast training

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Systems Engineering and Electronics 2024年第3期35卷 509-531,F0002页

作者： OU Xinwei CHEN Zhangxin ZHU Ce LIU Yipeng School of Information and Communication Engineering University of Electronic Science and Technology of ChinaChengdu 611731China

Deep neural networks(DNNs)have achieved great success in many data processing ***,high computational complexity and storage cost make deep learning difficult to be used on resource-constrained devices,and it is not environmental-friendly with much power *** this paper,we focus on low-rank optimization for efficient deep learning *** the space domain,DNNs are compressed by low rank approximation of the network parameters,which directly reduces the storage requirement with a smaller number of network *** the time domain,the network parameters can be trained in a few subspaces,which enables efficient training for fast *** model compression in the spatial domain is summarized into three categories as pre-train,pre-set,and compression-aware methods,*** a series of integrable techniques discussed,such as sparse pruning,quantization,and entropy coding,we can ensemble them in an integration framework with lower computational complexity and *** addition to summary of recent technical advances,we have two findings for motivating future *** is that the effective rank,derived from the Shannon entropy of the normalized singular values,outperforms other conventional sparse measures such as the?_1 norm for network *** other is a spatial and temporal balance for tensorized neural *** accelerating the training of tensorized neural networks,it is crucial to leverage redundancy for both model compression and subspace training.

关键词： model compression subspace training effective rank low rank tensor optimization efficient deep learning

Method for the automatic recognition of cropland headland images based on deep learning

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

International Journal of Agricultural and Biological Engineering 2023年第2期16卷 216-224页

作者： Yujie Qiao Hui Liu Zhijun Meng Jingping Chen Luyao Ma Information Engineering College Capital Normal UniversityBeijing 100048China National Engineering Research Center for Information Technology in Agriculture Beijing 100097China

For self-driving agricultural vehicles,the sensing of the headland environment based on image recognition is an important technological *** headland environments are complex and *** image feature extraction methods have many *** study proposed a method of automatic cropland headland image recognition based on deep *** on the characteristics of cropland headland environments and practical application needs,a dataset was constructed containing six categories of annotated cropland headland images and an augmented headland image training set was used to train the compact network *** the same experimental conditions,the model prediction accuracy for the first ranked category in all the results(Top-1 accuracy)of the MobileNetV2 network on the validation set was 98.5%.Compared with classic ResNetV2-50,Inception-V3,and backend-compressed Inception-V3,MobileNetV2 has a high accuracy,high recognition speed,and a small memory *** further test the performance of the model,250 images were used for each of the six categories of headland images as the test set for the *** average of the harmonic mean of precision and recall(F1-score)of the MobileNetV2 network for all the categories of headland images reached 97%.The MobileNetV2 network exhibits good robustness and *** results of this study indicate that onboard computers on self-driving agricultural vehicles are able to employ the MobileNetV2 network for headland image recognition to meet the application requirements of headland environment sensing.

关键词： cropland image deep learning image recognition model compression MobileNetV2 network

Detection of maize leaf diseases using improved MobileNet V3-small

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

International Journal of Agricultural and Biological Engineering 2023年第3期16卷 225-232页

作者： Ang Gao Aijun Geng Yuepeng Song Longlong Ren Yue Zhang Xiang Han College of Mechanical and Electrical Engineering Shandong Agricultural UniversityTai’an 271018ShandongChina Shandong Provincial Engineering Laboratory of Agricultural Equipment Intelligence Tai’an 271018ShandongChina Shandong Provincial Key Laboratory of Horticultural Machineries and Equipment Tai’an 271018ShandongChina

In order to realize the intelligent identification of maize leaf diseases for accurate prevention and control,this study proposed a maize disease detection method based on improved MobileNet V3-small,using a UAV to collect maize disease images and establish a maize disease dataset in a complex context,and explored the effects of data expansion and migration learning on model recognition accuracy,recall rate,and F1-score instructive evaluative indexes,and the results show that the two approaches of data expansion and migration learning effectively improved the accuracy of the *** structured compression of MobileNet V3-small bneck layer retains only 6 layers,the expansion multiplier of each layer was redesigned,32-fold fast downsampling was used in the first layer,and the location of the SE module was *** improved model had an average accuracy of 79.52%in the test set,a recall of 77.91%,an F1-score of 78.62%,a model size of 2.36 MB,and a single image detection speed of 9.02 *** detection accuracy and speed of the model can meet the requirements of mobile or embedded *** study provides technical support for realizing the intelligent detection of maize leaf diseases.

关键词： maize leaf disease image recognition model compression MobileNetV3-small

High-efficiency tea shoot detection method via a compressed deep learning model

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

International Journal of Agricultural and Biological Engineering 2022年第3期15卷 159-166,F0003页

作者： Yatao Li Leiying He Jiangming Jia Jianneng Chen Jun Lyu Chuanyu Wu Faculty of Mechanical Engineering&Automation Zhejiang Sci-Tech UniversityHangzhou 310018China Key Laboratory of Transplanting Equipment and Technology of Zhejiang Province Hangzhou 310018China School of Information Science and Technology Zhejiang Sci-Tech UniversityHangzhou 310018China

Achieving high-efficiency and accurate detection of tea shoots in fields are essential for tea robotic plucking. A real-time tea shoot detection method using the channel and layer pruned YOLOv3-SPP deep learning algorithm was proposed in this study. First, tea shoot images were collected and data augmentation was performed to increase sample diversity, and then a spatial pyramid pooling module was added to the YOLOv3 model to detect tea shoots. To simplify the tea shoot detection model and improve the detection speed, the channel pruning algorithm and layer pruning algorithm were used to compress the model. Finally, the model was fine-tuned to restore its accuracy, and achieve the fast and accurate detection of tea shoots. The test results demonstrated that the number of parameters, model size, and inference time of the tea shoot detection model after compression reduced by 96.82%, 96.81%, and 59.62%, respectively, whereas the mean average precision of the model was only 0.40% lower than that of the original model. In the field test, the compressed model was deployed on a Jetson Xavier NX to conduct the detection of tea shoots. The experimental results demonstrated that the detection speed of the compressed model was 15.9 fps, which was 3.18 times that of the original model. All the results indicate that the proposed method could be deployed on tea harvesting robots with low computing power to achieve high efficiency and accurate detection.

关键词： deep learning tea shoot detection model compression high-efficiency

Quant 4.0:engineering quantitative investmentwith automated,explainable,and knowledge-driven artificial intelligence

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Frontiers of Information Technology & Electronic Engineering 2024年第11期25卷 1421-1445页

作者： Jian GUO Saizhuo WANG Lionel M.NI Heung-Yeung SHUM IDEA Research International Digital Economy AcademyShenzhen 518045China The Hong Kong University of Science and Technology Hong Kong 999077China The Hong Kong University of Science and Technology(Guangzhou) Guangzhou 511453China

Quantitative investment(abbreviated as“quant”in this paper)is an interdisciplinary field combining financial engineering,computer science,mathematics,statistics,*** has become one of the mainstream investment methodologies over the past decades,and has experienced three generations:quant 1.0,trading by mathematical modeling to discover mis-priced assets in markets;quant 2.0,shifting the quant research pipeline from small“strategy workshops”to large“alpha factories”;quant 3.0,applying deep learning techniques to discover complex nonlinear pricing *** its advantage in prediction,deep learning relies on extremely large data volume and labor-intensive tuning of“black-box”neural network *** address these limitations,in this paper,we introduce quant 4.0 and provide an engineering perspective for next-generation *** 4.0 has three key differentiating ***,automated artificial intelligence(AI)changes the quant pipeline from traditional hand-crafted modeling to state-of-the-art automated modeling and employs the philosophy of“algorithm produces algorithm,model builds model,and eventually AI creates AI.”Second,explainable AI develops new techniques to better understand and interpret investment decisions made by machine learning black boxes,and explains complicated and hidden risk ***,knowledge-driven AI supplements data-driven AI such as deep learning and incorporates prior knowledge into modeling to improve investment decisions,in particular for quantitative value *** all these together,we discuss how to build a system that practices the quant 4.0 *** also discuss the application of large language models in quantitative ***,we propose 10 challenging research problems for quant technology,and discuss potential solutions,research directions,and future trends.

关键词： Artificial general intelligence Artificial intelligence Automated machine learning Causality engineering Deep learning Feature engineering Investment engineering Knowledge graph Knowledge reasoning Knowledge representation model compression Neural architecture search Quant 4.0 Quantitative investment Risk graph Explainable artificial intelligence

A Reinforcement Learning Training Acceleration Method Based on Knowledge Distillation

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

A Reinforcement Learning Training Acceleration Method Based ...