检索结果-内蒙古大学图书馆

Automatic diagnosis of disc herniation based on DenseNet fusion model 8

学校读者我要写书评

暂无评论

Automatic diagnosis of disc herniation based on DenseNet fus...

8th International Conference on Digital Home, ICDH 2020

作者： Wang, Zimin Qin, Junyan Huang, Jinrong Wang, Yumeng Li, Ji Guilin University of Electronic Technology Guangxi Colleges and Universities Key Lab. of Intelligent Processing of Computer Image and Graphics Guilin541004 China Guilin Huigu Institute of Artificial Intelligence Industrial Technology Guilin541004 China Guilin University of Electronic Technology School of Computer Science and Information Security Guilin541004 China

ISBN: (纸本)9781728192345

Spine MRI images generally have the characteristics of low contrast and much noise. Because the variable shape of the spine edge, the traditional spine image segmentation method requires a lot of preprocessing and can not perform image segmentation automatically, which is inefficient and cannot compliant the clinical work requirements. So as to solve the above problems, this paper use deep learning algorithms to identify the spine images, and proposed a disc herniation recognition method for spinal MRI images which based on DenseNet fusion model. Use DenseNet and Inception-Resnet-V2 model fusion to get more image features and improve the image recognition accuracy. Experiments prove that this method can be used to identify the herniation of spinal intervertebral disc with high accuracy and real-time. © 2020 IEEE.

关键词： image segmentation

A Novel LOS Rate Estimation Method Based on images for Strap-down Inertial Guidance 3

学校读者我要写书评

暂无评论

A Novel LOS Rate Estimation Method Based on Images for Strap...

2020 3rd International Conference on Advanced Algorithms and Control Engineering, ICAACE 2020

作者： Xu, Zheng Luo, Haibo Hui, Bin Chang, Zheng Shenyang Institute of Automation Chinese Academy of Sciences Shenyang110016 China Institutes for Robotics and Intelligent Manufacturing Chinese Academy of Sciences Shenyang110016 China University of Chinese Academy of Sciences Beijing100049 China Key Laboratory of Opto-Electronic Information Processing Chinese Academy of Science Shenyang110016 China Key Lab of Image Understanding and Computer Vision Shenyang110016 China

With the development of technology, precision guided weapon is becoming more and more important in modern war. In order to launch our recent guidance system on medium and small guided weapons, we propose a method to obtain the LOS rate by combining information from both camera and gyroscope. To be specific, we firstly calculate the body LOS angle through transforming the image pixel coordinate system into the image physical coordinate system according to camera internal parameters;then subtract the missile motion information contained in the measurement signal of the seeker and finally the LOS rate is deduced. Comparing with traditional gimballed seekers, our strap-down seekers with camera have significantly reduced costs and influences caused by external environments on platform. © Published under licence by IOP Publishing Ltd.

关键词： Cameras

ER-IQA: Boosting perceptual quality assessment using external reference images

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Guo, Jingyu Wang, Wei Yang, Wenming Liao, Qingmin Zhou, Jie The Shenzhen Key Laboratory of Visual Image Processing Shenzhen Internatinal Graduate School Tsinghua University Shenzhen518055 China The Department of Electronic Engineering Tsinghua University Beijing100084 China The AI Lab ByteDance Beijing100000 China The Department of Automation Tsinghua University Beijing100084 China

Recently, image quality assessment (IQA) has achieved remarkable progress with the success of deep learning. However, the strict pre-condition of full-reference (FR) methods has limited its application in real scenarios. And the no-reference (NR) scheme is also inconvenient due to its unsatisfying performance as a result of ignoring the essence of image quality. In this paper, we introduce a brand new scheme, namely external-reference image quality assessment (ER-IQA), by introducing external reference images to bridge the gap between FR and NR-IQA. As the first implementation and a new baseline of ER-IQA, we propose a new Unpaired-IQA network to process images in an content-unpaired manner. A Mutual Attention-based Feature Enhancement (MAFE) module is well-designed for the unpaired features in ER-IQA. The MAFE module allows the network to extract quality-discriminative features from distorted images and content variability-robust features from external reference ones. Extensive experiments demonstrate that the proposed model outperforms the state-of-the-art NR-IQA methods, verifying the effectiveness of ER-IQA and the possibility of narrowing the gap of the two existing categories. Copyright © 2021, The Authors. All rights reserved.

关键词： image quality

Exploiting learnable joint groups for hand pose estimation

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Li, Moran Gao, Yuan Sang, Nong Key Laboratory of Image Processing and Intelligent Control School of Artificial Intelligence and Automation Huazhong University of Science and Technology Wuhan China Tencent AI Lab United States

In this paper, we propose to estimate 3D hand pose by recovering the 3D coordinates of joints in a group-wise manner, where less-related joints are automatically categorized into different groups and exhibit different features. This is different from the previous methods where all the joints are considered holistically and share the same feature. The benefits of our method are illustrated by the principle of multi-task learning (MTL), i.e., by separating less-related joints into different groups (as different tasks), our method learns different features for each of them, therefore efficiently avoids the negative transfer (among less related tasks/groups of joints). The key of our method is a novel binary selector that automatically selects related joints into the same group. We implement such a selector with binary values stochastically sampled from a Concrete distribution, which is constructed using Gumbel softmax on trainable parameters. This enables us to preserve the differentiable property of the whole network. We further exploit features from those less-related groups by carrying out an additional feature fusing scheme among them, to learn more discriminative features. This is realized by implementing multiple 1x1 convolutions on the concatenated features, where each joint group contains a unique 1x1 convolution for feature fusion. The detailed ablation analysis and the extensive experiments on several benchmark datasets demonstrate the promising performance of the proposed method over the state-of-the-art (SOTA) methods. Besides, our method achieves top-1 among all the methods that do not exploit the dense 3D shape labels on the most recently released FreiHAND competition at the submission date. The source code and models are available at https://***/moranli-aca/LearnableGroups-Hand. Copyright © 2020, The Authors. All rights reserved.

关键词： Convolution

A Novel Triboelectric-Optical Hybrid Tactile Sensor for Human-Machine Tactile Interaction

学校读者我要写书评

暂无评论

SSRN

SSRN 2023年

作者： Yang, Hui Bu, Tianzhao Liu, Wenbo Liu, Jiaqi Ling, Yunzhi Wu, Meixia Liu, Weirui Wang, Changan Gao, Xifeng Wang, Lihui Dynamic Image Perception Lab Institute of Semiconductors Guangdong Academy of Sciences Guangdong510075 China Biomechanics and Soft Robotics Lab School of Mechanical Engineering and Automation Beihang University Beijing100083 China School of Mechanical Science and Engineering Huazhong University of Science & Technology Wuhan430074 China School of Mechanical Engineering and Automation Liaoning Petrochemical University Fushun113001 China School of Electronics & Communication Guangdong Mechanical and Electrical Polytechnic Guangzhou510515 China Center of Ultra-Precision Optoelectronic Instrumentation Engineering Harbin Institute of Technology Harbin150001 China Key Lab of Ultra-Precision Intelligent Instrumentation Ministry of Industry Information Technology Harbin150080 China

Tactile interaction is one of the primary modes of human-machine interaction, which requires artificial devices to possess tactile sensing abilities. In recent years, to enrich the tactile information perceived by artificial devices, tactile sensors based on hybrid sensing mechanisms have attracted extensive attention. However, for wearable or flexible artificial devices, the development of tactile sensors with a simple structure and preparation process, low cost and power consumption, and good maintainability and integration remains a tough challenge. Herein, we presented a triboelectric-optical hybrid tactile sensor (TOHTS) with a separable modular structure, comprising two units: a triboelectric unit designed based on the single-electrode triboelectric nanogenerator to receive external tactile stimuli, and a photoelectric unit with a liquid-membrane liquid lens structure and a function of translating the tactile stimuli into a photoelectric signal. Both units exhibit a simple structure, and their parts can be easily cannibalized and replaced. The triboelectric unit can control the on-off behavior of the light source embedded in the photoelectric unit to minimize power consumption and improve tactile response speed, according to its output signal. The photoelectric unit possesses superior electromagnetic interference immunity, making the TOHTS accurately and quantitatively perceive contact force. The sensing performance test results validate the TOHTS’s advantages of a short response time (about 9 ms), strong linearity (R2≈0.9952) in sensing contact force, and excellent durability and stability. Due to the above features, the TOHTS exhibits outstanding practicability in tactile interactive tasks, which is proved by some interactive input experiments, and it will have extensive application prospects for human-machine interactive devices. © 2023, The Authors. All rights reserved.

关键词： Nanogenerators

Evolutionary neural architecture search for retinal vessel segmentation

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Fan, Zhun Wei, Jiahong Zhu, Guijie Mo, Jiajie Li, Wenji Department of Electronic and Information Engineering Shantou University Shantou515063 China Key Lab of Digital Signal and Image Processing of Guangdong Province Shantou University Shantou515063 China

The accurate retinal vessel segmentation (RVS) is of great significance to assist doctors in the diagnosis of ophthalmology diseases and other systemic diseases. Manually designing a valid neural network architecture for retinal vessel segmentation requires high expertise and a large workload. In order to improve the performance of vessel segmentation and reduce the workload of manually designing neural network, we propose novel approach which applies neural architecture search (NAS) to optimize an encoder-decoder architecture for retinal vessel segmentation. A modified evolutionary algorithm is used to evolve the architectures of encoder-decoder framework with limited computing resources. The evolved model obtained by the proposed approach achieves top performance among all compared methods on the three datasets, namely DRIVE, STARE and CHASE_DB1, but with much fewer parameters. Moreover, the results of cross-training show that the evolved model is with considerable scalability, which indicates a great potential for clinical disease diagnosis. Copyright © 2020, The Authors. All rights reserved.

关键词： Ophthalmology

Alternate INRAE-Bordeaux Soil Moisture and L-Band Vegetation Optical Depth Products from SMOS and SMAP: Current Status and Overview

学校读者我要写书评

暂无评论

Alternate INRAE-Bordeaux Soil Moisture and L-Band Vegetation...

IEEE International Symposium on Geoscience and Remote Sensing (IGARSS)

作者： Xiaojun Li Roberto Fernandez-Moran F. Frappart Lei Fan Gabrielle De Lannoy Xiangzhuo Liu Huan Wang Zanping Xing Mengjia Wang Yao Xiao C. Moisy Jean-Pierre Wigneron INRAE UMR1391 ISPA Villenave d'Ornon France Image Processing Lab (IPL) University of Valencia Valencia Spain Chongqing Jinfo Mountain Karst Ecosystem National Observation and Research Station School of Geographical Sciences Southwest University Chongqing China Department of Earth and Environmental Sciences KU Leuven Heverlee Belgium College of Urban and Environmental Sciences Key Laboratory for Earth Surface Processes of the Ministry of Education Peking University Beijing China Cryosphere Research Station on the Qinghai-Tibet Plateau State Key Laboratory of Cryospheric Science Northwest Institute of Eco-Environment and Resource Chinese Academy of Sciences Lanzhou China School of Geoscience and Technology Zhengzhou University China

Between 2018 and 2022, INRAE Bordeaux (IB) has developed a series of soil moisture (SM) and L-band Vegetation Optical depth (L-VOD) retrieval products from SMOS and SMAP, which are currently the only two operational L-band passive microwave satellite missions. These IB products rely on a two-parameter inversion of the L-MEB model (L-band Microwave Emission of the Biosphere) which requires little ancillary information. The products are found to be accurate, and very well-suited for application in hydrology, agriculture, climate and vegetation monitoring. In this communication, we present an overview of the development, evaluation and new applications of these IB SM or L-VOD products.

关键词：

Frequency-Aware Physics-Inspired Degradation Model for Real-World image Super-Resolution

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Dong, Zhenxing Cao, Hong Shen, Wang Gan, Yu Ling, Yuye Zhai, Guangtao Su, Yikai John Hopcroft Center for Computer Science Shanghai Jiao Tong University China Institute of Image Communication and Network Engineering Shanghai Jiao Tong University China Department of Electrical and Computer Engineering University of Alabama TuscaloosaAL35401 United States State Key Lab of Advanced Optical Communication Systems and Networks Shanghai Jiao Tong University China

Current learning-based single image super-resolution (SISR) algorithms underperform on real data due to the deviation in the assumed degradation process from that in the real-world scenario. Conventional degradation processes consider applying blur, noise, and downsampling (typically bicubic downsampling) on high-resolution (HR) images to synthesize low-resolution (LR) counterparts. However, few works on degradation modelling have taken the physical aspects of the optical imaging system into consideration. In this paper, we analyze the imaging system optically and exploit the characteristics of the real-world LR-HR pairs in the spatial frequency domain. We formulate a real-world physics-inspired degradation model by considering both optics and sensor degradation;The physical degradation of an imaging system is modelled as a low-pass filter, whose cut-off frequency is dictated by the object distance, the focal length of the lens, and the pixel size of the image sensor. In particular, we propose to use a convolutional neural network (CNN) to learn the cutoff frequency of real-world degradation process. The learned network is then applied to synthesize LR images from unpaired HR images. The synthetic HR-LR image pairs are later used to train an SISR network. We evaluate the effectiveness and generalization capability of the proposed degradation model on real-world images captured by different imaging systems. Experimental results showcase that the SISR network trained by using our synthetic data performs favorably against the network using the traditional degradation model. Moreover, our results are comparable to that obtained by the same network trained by using real-world LR-HR pairs, which are challenging to obtain in real scenes. © 2021, CC BY.

关键词： Cutoff frequency

A multi-user oriented live free-viewpoint video streaming system based on view interpolation

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Hu, Jingchuan Guo, Shuai Dong, Yu Zhou, Kai Xu, Jun Song, Li Institute of Image Communication and Network Engineering Shanghai Jiao Tong University Shanghai200240 China Cooperative Medianet Innovation Center Shanghai Jiao Tong University Shanghai200240 China MoE Key Lab of Artificial Intelligence AI Institute Shanghai Jiao Tong University Shanghai200240 China

As an important application form of immersive multimedia services, free-viewpoint video(FVV) enables users with great immersive experience by strong interaction. However, the computational complexity of virtual view synthesis algorithms poses a significant challenge to the real-time performance of an FVV system. Furthermore, the individuality of user interaction makes it difficult to serve multiple users simultaneously for a system with conventional architecture. In this paper, we novelly introduce a CNN-based view interpolation algorithm to synthesis dense virtual views in real time. Based on this, we also build an end-to-end live free-viewpoint system with a multi-user oriented streaming strategy. Our system can utilize a single edge server to serve multiple users at the same time without having to bring a large view synthesis load on the client side. We analysis the whole system and show that our approaches give the user a pleasant immersive experience, in terms of both visual quality and latency. Copyright © 2021, The Authors. All rights reserved.

关键词： Multimedia services