Filters








384 Hits in 2.9 sec

Quadtree Convolutional Neural Networks [chapter]

Pradeep Kumar Jayaraman, Jianhan Mei, Jianfei Cai, Jianmin Zheng
2018 Lecture Notes in Computer Science  
This paper presents a Quadtree Convolutional Neural Network (QCNN) for efficiently learning from image datasets representing sparse data such as handwriting, pen strokes, freehand sketches, etc.  ...  The actual image data corresponding to non-zero pixels is stored in the finest nodes of the quadtree.  ...  Quadtree Convolution Neural Network Motivation Consider a general scenario where a dense n-dimensional tensor used to represent some input that is to be fed into a convolutional neural network.  ... 
doi:10.1007/978-3-030-01231-1_34 fatcat:gynykn7lc5bo5pntbs2qcgi34m

Grain depot image dehazing via quadtree decomposition and convolutional neural networks

Zhihui Li, Bian Gui, Tong Zhen, Yuhua Zhu
2020 Alexandria Engineering Journal  
Li et al., Grain depot image dehazing via quadtree decomposition and convolutional neural networks, Alexandria Eng. J. (2020), https://doi.  ...  after a series of feature learning of the network, and then the image fusion method is used to refine it.  ...  Li et al., Grain depot image dehazing via quadtree decomposition and convolutional neural networks, Alexandria Eng. J. (2020), https://doi.org/10.1016/j.aej.2020.03.048  ... 
doi:10.1016/j.aej.2020.03.048 fatcat:iht3rk2owzd2ndktwxl5lt34hq

Quadtree Generating Networks: Efficient Hierarchical Scene Parsing with Sparse Convolutions [article]

Kashyap Chitta, Jose M. Alvarez, Martial Hebert
2019 arXiv   pre-print
Semantic segmentation with Convolutional Neural Networks is a memory-intensive task due to the high spatial resolution of feature maps and output predictions.  ...  In this paper, we present Quadtree Generating Networks (QGNs), a novel approach able to drastically reduce the memory footprint of modern semantic segmentation networks.  ...  Fully Convolutional Networks (FCNs) [25] and their variants [4, 27, 33, 46] have become a standard tool for this task, leveraging the representational power of deep Convolutional Neural Networks (CNNs  ... 
arXiv:1907.11821v2 fatcat:absnoy5i5zeiva26nzjbgbpoju

Cloud Detection Method Using CNN Based on Cascaded Feature Attention and Channel Attention

Jing Zhang, Jun Wu, Hui Wang, Yuchen Wang, Yunsong Li
2021 IEEE Transactions on Geoscience and Remote Sensing  
In this paper, we propose a cloud detection method using convolutional neural network based on cascaded feature attention and channel attention (CFCA-Net).  ...  Moreover, a loss function combined quadtree and binary cross-entropy was also introduced to make the network focus on the edge of cloud area.  ...  The Fully Convolutional Network (FCN) [36] introduced an endto-end fully convolutional neural network structure for semantic segmentation.  ... 
doi:10.1109/tgrs.2021.3120752 fatcat:2vhyjz5yireqpd4yszva7js7q4

Learning Sparse Feature Representations Using Probabilistic Quadtrees and Deep Belief Nets

Saikat Basu, Manohar Karki, Sangram Ganguly, Robert DiBiano, Supratik Mukhopadhyay, Shreekant Gayaka, Rajgopal Kannan, Ramakrishna Nemani
2016 Neural Processing Letters  
On the MNIST and n-MNIST datasets, our framework shows promising results and significantly outperforms traditional Deep Belief Networks.  ...  Then, we propose a novel framework for the classification of handwritten digits that learns sparse representations using probabilistic quadtrees and Deep Belief Nets.  ...  learning phase using Feedforward Backpropagation Neural Networks.  ... 
doi:10.1007/s11063-016-9556-4 fatcat:pxei5ue72nasloobtcb43irjla

Pixel-level Reconstruction and Classification for Noisy Handwritten Bangla Characters [article]

Manohar Karki, Qun Liu, Robert DiBiano, Saikat Basu, Supratik Mukhopadhyay
2018 arXiv   pre-print
Quadtrees can be an efficient representation for learning from sparse features.  ...  The pixel level denoiser (a deep belief network) uses the map responses obtained from a pretrained CNN as features for reconstructing the characters eliminating noise.  ...  Map responses from intermediate layers of a Convolutional Neural Network have been used as features in [17] to segment images.  ... 
arXiv:1806.08037v1 fatcat:yde7metennao5p5xvo44vtr54e

Guest Editorial Skin Lesion Image Analysis for Melanoma Detection

M. Emre Celebi, Noel Codella, Allan Halpern, Dinggang Shen
2019 IEEE journal of biomedical and health informatics  
In Improving Dermoscopic Image Segmentation with Enhanced Convolutional-Deconvolutional Networks, Yuan and Lo propose a convolutional-deconvolutional neural network (CDNN) architecture for segmentation  ...  Finally, in Fully Convolutional Neural Networks to Detect Clinical Dermoscopic Features, Kawahara and Hamarneh describe their system that won the 2017 ISIC Challenge for Part 2: Feature Classification  ... 
doi:10.1109/jbhi.2019.2897338 fatcat:3cszhtjw6zc5thl4abzeaukq3q

Learning Sparse Feature Representations using Probabilistic Quadtrees and Deep Belief Nets [article]

Saikat Basu, Manohar Karki, Sangram Ganguly, Robert DiBiano, Supratik Mukhopadhyay, Ramakrishna Nemani
2015 arXiv   pre-print
On the MNIST and n-MNIST datasets, our framework shows promising results and significantly outperforms traditional Deep Belief Networks.  ...  Then, we propose a novel framework for the classification of handwritten digits that learns sparse representations using probabilistic quadtrees and Deep Belief Nets.  ...  learning phase using Feedforward Backpropagation Neural Networks.  ... 
arXiv:1509.03413v1 fatcat:fcvgebl3cffzlhiuckvyuuxmwi

A QuadTree Image Representation for Computational Pathology [article]

Rob Jewsbury, Abhir Bhalerao, Nasir Rajpoot
2021 arXiv   pre-print
Histopathology images are large and need to be split up into image tiles or patches so modern convolutional neural networks (CNNs) can process them.  ...  To the best of our knowledge, this is the first attempt to use quadtrees for pathology image data.  ...  [11] use an Octree-based Convolutional Neural Network (O-CNN) for 3D shape analysis.  ... 
arXiv:2108.10873v1 fatcat:v5fknwvb5nexficcfrbh5tlix4

Semantic Segmentation of Remote Sensing Images Combining Hierarchical Probabilistic Graphical Models and Deep Convolutional Neural Networks

Martina Pastorino, Gabriele Moser, Sebastiano B. Serpico, Josiane Zerubia
2021 2021 IEEE International Geoscience and Remote Sensing Symposium IGARSS  
Recent advances in deep learning (DL), especially convolutional neural networks (CNNs) and fully convolutional networks (FCNs), have shown outstanding performances in this task.  ...  The considered architectures have 5 convolutional blocks, each containing convolutional layers, zero paddings, followed by ReLU activations and batch normalizations.  ...  Three skip connections, from three deconvolution blocks of the decoder to the output layer allow to collect the activations of the network at different resolutions, inserted into the quadtree to connect  ... 
doi:10.1109/igarss47720.2021.9553253 fatcat:k2bw5nyfznbx7eqfqmwwnv5i5m

Segmentation Techniques for Medical Images – An Appraisal

S. Rakoth, J. Sasikala
2016 International Journal of Computer Applications  
In order to classify the segmentation techniques such as GA, Neural Network, Soft Computing and various image segmentation techniques and their performances analysis is done.  ...  Then the new images were segmented with trained neural network. Neural network segmentation method includes two important steps: feature extraction and image segmentation based on neural network.  ...  Neural Network Approach Neural Network is formed by several elements that are connected by various links with variable weights [17] .  ... 
doi:10.5120/ijca2016912174 fatcat:lum2xqt4y5eexbtczgqgccmw2q

Real-Time Semantic Segmentation with Label Propagation [chapter]

Rasha Sheikh, Martin Garbade, Juergen Gall
2016 Lecture Notes in Computer Science  
Despite of the success of convolutional neural networks for semantic image segmentation, CNNs cannot be used for many applications due to limited computational resources.  ...  Introduction Although convolutional neural networks have shown a great success for semantic image segmentation in the last years [1] [2] [3] , fast inference can only be achieved by massive parallelism  ...  In the last years, convolutional neural networks have become very popular for semantic segmentation [1, 2, 10] . Recent approaches achieve accurate segmentation results even without CRFs [3] .  ... 
doi:10.1007/978-3-319-48881-3_1 fatcat:llw3khvepjhnxkict7mpa3khn4

Removing Blocking Artifacts in Video Streams Using Event Cameras [article]

Henry H. Chopp, Srutarshi Banerjee, Oliver Cossairt, Aggelos K. Katsaggelos
2021 arXiv   pre-print
In this paper, we propose EveRestNet, a convolutional neural network designed to remove blocking artifacts in videostreams using events from neuromorphic sensors.  ...  We first degrade the video frame using a quadtree structure to produce the blocking artifacts to simulate transmitting a video under a heavily constrained bandwidth.  ...  However, EveRestNet is convolutional, so all dimensions need to be discretized in order to be valid inputs into the neural network.  ... 
arXiv:2105.05973v1 fatcat:gick426ysnhorpjnyteuunimiq

Image Classification using Graph Neural Network and Multiscale Wavelet Superpixels [article]

Varun Vasudevan and Maxime Bassenne and Md Tauhidul Islam and Lei Xing
2022 arXiv   pre-print
Prior studies using graph neural networks (GNNs) for image classification have focused on graphs generated from a regular grid of pixels or similar-sized superpixels.  ...  We use SplineCNN, a state-of-the-art network for image graph classification, to compare WaveMesh and similar-sized superpixels.  ...  Introduction Convolutional neural networks (CNNs) achieve the best performance on various image classification tasks.  ... 
arXiv:2201.12633v1 fatcat:up2pjtg7gvdq3c4rdiub4mm6x4

Enhancement of Tuberculosis Detection Using Ensemble Classifier with Quadtree Method: A Preliminary Study

Laura Jack, UXRL, Faculty of Computing and Informatics, Universiti Malaysia Sabah, Malaysia
2020 Journal of Modern Mechanical Engineering and Technology  
Convolutional Neural Network (CNN) is another approach to analyze images and capture the important information of images and use it to classify correctly input images into respective categories.  ...  Second classification method will used Convolutional Neural Network (CNN) as classifier and third method will use Ensemble classifier on which is the voting method on first and second classification method  ... 
doi:10.31875/2409-9848.2020.07.1 fatcat:5vb45pf4rfblbkjpx3gkon5uny
« Previous Showing results 1 — 15 out of 384 results