1,829 Hits in 3.7 sec

Deep Controllable Backlight Dimming [article]

Lvyin Duan, Demetris Marnerides, Alan Chalmers, Zhichun Lei, Kurt Debattista
2020 arXiv   pre-print
In this work, a novel deep learning based local dimming method is proposed for rendering HDR images on dual-panel HDR displays.  ...  Dual-panel displays require local dimming algorithms in order to reproduce content with high fidelity and high dynamic range.  ...  In this paper, a novel local dimming algorithm based on a CNN architecture is proposed for displaying HDR images on dual-panel HDR monitors.  ... 
arXiv:2008.08352v1 fatcat:7jvyiyuijzernbqjk5xi6rl2eu

The Enhancement of WiFi Fingerprint Positioning Using Convolutional Neural Network

2018 DEStech Transactions on Computer Science and Engineering  
In this paper, we proposed a new CNN (convolutional neural network)-based architecture for WiFi fingerprint positioning.  ...  With the widely applied of LBS (Location Based Service), a high standard of indoor positioning technology is required. Among them, WiFi fingerprint positioning is the most popular way.  ...  Model precision CNN-based(dim = 10) 90.49% CNN-based(dim = 20) 90.71% CNN-based(dim = 50) 90.83% KNN (k = 3) 85.88% KNN (k = 5) 86.12% KNN (k = 10) 85.91%  ... 
doi:10.12783/dtcse/ccnt2018/24745 fatcat:3gzfjadn3nhp5p3tyrgifrfnva

Hand PointNet: 3D Hand Pose Estimation Using Point Sets

Liuhao Ge, Yujun Cai, Junwu Weng, Junsong Yuan
2018 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition  
Different from existing CNN-based hand pose estimation methods that take either 2D images or 3D volumes as the input, our proposed Hand PointNet directly processes the 3D point cloud that models the visible  ...  Convolutional Neural Network (CNN) has shown promising results for 3D hand pose estimation in depth images.  ...  This figure is best viewed in color. back loop (Feedback) [21] , 2D CNN for hand model regression (DeepModel) [50] , matrix completion with deep feature (DeepHand) [29] , local surface normal based  ... 
doi:10.1109/cvpr.2018.00878 dblp:conf/cvpr/GeCWY18 fatcat:z7zku3vbj5ejtpwizlnrnl52s4

Research on Anti-Alzheimer's Traditional Chinese Medicine with Data Security: Datasets, Methods, and Evaluation

Peng Tang, Yubin Zheng, Weidong Qiu, Hao Wang, Jie Guo, Zheng Huang, Zhe-Li Liu
2022 Security and Communication Networks  
Finally, we built a distributed model training architecture based on federated learning to train and predict the effectiveness of TCM prescriptions under the premise of ensuring data security.  ...  Although deep learning technologies such as MLP and CNN have been widely used in various fields, machine learning technologies such as SVM will also achieve better results when the amount of data is small  ...  With the rise of big data analysis and deep learning technology, a new path has been opened up for the development of TCM.  ... 
doi:10.1155/2022/7697863 fatcat:fqqy5hii2ngstdaz6n2actprha

Actigraphy-based Sleep/Wake Pattern Detection using Convolutional Neural Networks [article]

Lena Granovsky, Gabi Shalev, Nancy Yacovzada, Yotam Frank, Shai Fine
2018 arXiv   pre-print
The proposed methods are a sequential CNN, reminiscent of the bi-directional CNN for slot filling, and a Multi-Task Learning (MTL) based model.  ...  This study presents two novel modeling schemes that utilize Deep Convolutional Neural Networks (CNN) to identify sleep/wake states.  ...  In the sequel, we will present two 1-Dim CNN-based models.  ... 
arXiv:1802.07945v1 fatcat:fxxdsnus5jhfzkhf57wdrm6n2m

Automatic Gaze Analysis: A Survey of Deep Learning based Approaches [article]

Shreya Ghosh, Abhinav Dhall, Munawar Hayat, Jarrod Knibbe, Qiang Ji
2021 arXiv   pre-print
We analyze recent gaze estimation and segmentation methods, especially in the unsupervised and weakly supervised domain, based on their advantages and reported evaluation metrics.  ...  regression forest [118] , deep learning based landmark localization models [96] , [105] , heterogeneous CNN models [119] etc.  ...  The heuristic based approaches mainly include motion localization [128] and template matching [129] .  ... 
arXiv:2108.05479v2 fatcat:o5v4ueaqh5es5pogvagp2y3vv4

A Novel Temporal Attentive-Pooling based Convolutional Recurrent Architecture for Acoustic Signal Enhancement [article]

Tassadaq Hussain, Wei-Chien Wang, Mandar Gogate, Kia Dashtipour, Yu Tsao, Xugang Lu, Adeel Ahsan, Amir Hussain
2022 arXiv   pre-print
The existing deep acoustic signal enhancement (ASE) architectures ignore this kind of local and global effect.  ...  The proposed approach considers both global and local attention for ASE tasks.  ...  𝑑𝑖𝑚 , and 𝑾 𝑟 ∈ 𝑅 𝑁 𝑟 ×𝑟𝑛𝑛 𝑑𝑖𝑚 are the parameter matrices used to concatenate 𝒚(𝑡) and 𝒉(𝑇) , where 𝑐𝑛𝑛 𝑑𝑖𝑚 and 𝑟𝑛𝑛 𝑑𝑖𝑚 are the dimensions of the convolutional and LSTM layer  ... 
arXiv:2201.09913v1 fatcat:nromw5clt5dyxg6gbjtfh2ef4e

AI Oriented Large-Scale Video Management for Smart City: Technologies, Standards and Beyond [article]

Lingyu Duan, Yihang Lou, Shiqi Wang, Wen Gao, Yong Rui
2017 arXiv   pre-print
Deep learning has achieved substantial success in a series of tasks in computer vision.  ...  To enable interoperability in the context of deep feature coding, standardization is urgent and important.  ...  CNN-based retrieval methods can be categorized into two types: pre-trained and fine-tuned CNN models.  ... 
arXiv:1712.01432v1 fatcat:7ollwfwufzbpfljdgqp2lodfcm

Assistive Vision Technology using Deep Learning Techniques

Dr. Neeta Verma
2021 International Journal for Research in Applied Science and Engineering Technology  
In our model, first task is to detect objects within the image using pre trained Mobilenet model of CNN (Convolutional Neural Networks) and therefore the other is to caption the pictures based on the detected  ...  The model uses pre-trained CNN and LSTM models to perform the task of extracting objects or features to get the captions.  ...  CNN-based Image Feature Extractor For feature extraction, we use a CNN.  ... 
doi:10.22214/ijraset.2021.36815 fatcat:rgrjpywcivcmpfwhq5gnyhobay

CNN Approaches for Classification of Indian Leaf Species Using Smartphones

M. Vilasini, P. Ramamoorthy
2019 Computers Materials & Continua  
This paper discusses CNN based approaches for Indian leaf species identification from white background using smartphones.  ...  Variations of CNN models over the features like traditional shape, texture, color and venation apart from the other miniature features of uniformity of edge patterns, leaf tip, margin and other statistical  ...  CNN 2 71.9% 5 56.2% 10 43.9% Table 4 : 4 Results of CNN (softmax) for various dataset labels Dataset Labels CNN 2 100% 5 92% 10 88% KPR Institute of Engineering and Technology  ... 
doi:10.32604/cmc.2020.08857 fatcat:aigspccjaje3bprvkmviusf2bq

Incorporating Noise Robustness in Speech Command Recognition by Noise Augmentation of Training Data

Ayesha Pervaiz, Fawad Hussain, Huma Israr, Muhammad Ali Tahir, Fawad Riasat Raja, Naveed Khan Baloch, Farruh Ishmanov, Yousaf Bin Zikria
2020 Sensors  
We thoroughly analyse the latest trends in speech recognition and evaluate the speech command dataset on different machine learning based and deep learning based techniques.  ...  The advent of new devices, technology, machine learning techniques, and the availability of free large speech corpora results in rapid and accurate speech recognition.  ...  Acknowledgments: This research was supported by Yeungnam University, Korea, Kwangwoon University, Korea, and the University of Engineering and Technology, Taxila, Pakistan.  ... 
doi:10.3390/s20082326 pmid:32325814 pmcid:PMC7219662 fatcat:ftbpxexwd5fvbpj4s2cr76uybq


Y. Wang, D. Gong, H. Hu, S. Wang, Y. Han, Y. Wang, X. Ma
2021 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences  
On Middlebury and KITTI datasets, learning-based algorithms has shown their superiority compared to SGM derived methods. In this context, we assume that matching cost is the key factor of DIM.  ...  This paper reviews and evaluates Census Transform, and MC-CNN on a WorldView-3 typical city scene satellite stereo images on the premise that the overall SGM framework remains unchanged, providing a preliminary  ...  Matching Cost-CNN Dense matching based on deep learning is generally divided into two strategies: learning only part of the four steps of classic dense matching, namely non-end-to-end learning and end-to-end  ... 
doi:10.5194/isprs-archives-xliii-b2-2021-109-2021 fatcat:ktxw4kuytvez5lou6hwiidghgu

Predicting Sentiment Polarity of Microblogs using an LSTM – CNN Deep Learning Model

2019 International Journal of Engineering and Advanced Technology  
It was observed that the proposed model performs better and gives improved prediction accuracy when compared to semantic, machine learning and deep neural network approaches such as SVM, CNN, LSTM, CNN-LSTM  ...  The encodings produced by the LSTM layer are then fed to a CNN layer, which generates localized patterns of higher accuracy.  ...  : 64 -- Activation: ReLu Filter Length: 3 Output Dim: 64 Activation: ReLu Filter Length: 3 Output Dim: 64 VI.  ... 
doi:10.35940/ijeat.f8933.088619 fatcat:uo5golwcy5dibon5hx53mde6ni

Video-Based Parking Occupancy Detection for Smart Control System

Chen, Sheu, Peng, Wu, Tseng
2020 Applied Sciences  
The proposed method adopts You Only Look Once version 3 (YOLO v3, Seattle, WA, USA) based on MobileNet version 2 (MobileNet v2, Salt Lake City, UT, USA), which is area-based and uses voting to stably recognize  ...  However, traditional parking occupancy systems are mostly implemented for outdoor environments using costly sensor-based techniques.  ...  Therefore, our novel approach was proposed for existing streetlight cameras that use technology with deep learning and match with local computers to calculate occupancy results for parking spaces on entire  ... 
doi:10.3390/app10031079 fatcat:c22sc2kdojblrbavz6gozqq2ia

An efficient multiclassifier system based on convolutional neural network for offline handwritten Telugu character recognition

Soumya T Soman, Ashakranthi Nandigam, V. Srinivasa Chakravarthy
2013 2013 National Conference on Communications (NCC)  
The four techniques are: 1) Convolutional neural networks (CNN), 2) Principal Component Analysis (PCA), 3) Support vector machines, 4) Multiclassifier systems.  ...  Hybrid approach for CNN training: Gradient descent-based classical training algorithms applied to CNNs are very slow.  ...  These features make CNNs most attractive for image pattern recognition. But being large multilayer network architectures, CNNs suffer from the disadvantage of slow training and local minima. B.  ... 
doi:10.1109/ncc.2013.6488008 fatcat:2262cgs6andtth47vvkakpwizq
« Previous Showing results 1 — 15 out of 1,829 results