14,669 Hits in 2.8 sec

Transform-Invariant Convolutional Neural Networks for Image Classification and Search [article]

Xu Shen, Xinmei Tian, Anfeng He, Shaoyan Sun, Dacheng Tao
2019 arXiv   pre-print
Convolutional neural networks (CNNs) have achieved state-of-the-art results on many visual recognition tasks.  ...  Rather, each convolutional kernel learns to detect a feature that is generally helpful for producing the transform-invariant answer given the combinatorially large variety of transform levels of its input  ...  TRANSFORM-INVARIANT CONVOLU-TIONAL NEURAL NETWORKS Convolutional Neural Networks Convolutional neural networks (CNNs) are a supervised feed-forward multi-layer architecture in which each layer learns  ... 
arXiv:1912.01447v1 fatcat:yoa2syxqcjgqpbd4psmk4ntxeq

Transform-Invariant Convolutional Neural Networks for Image Classification and Search

Xu Shen, Xinmei Tian, Anfeng He, Shaoyan Sun, Dacheng Tao
2016 Proceedings of the 2016 ACM on Multimedia Conference - MM '16  
Convolutional neural networks (CNNs) have achieved stateof-the-art results on many visual recognition tasks.  ...  from transformed input images to transform-invariant representations.  ...  TRANSFORM-INVARIANT CONVOLU-TIONAL NEURAL NETWORKS Convolutional Neural Networks Convolutional neural networks (CNNs) are a supervised feed-forward multi-layer architecture in which each layer learns  ... 
doi:10.1145/2964284.2964316 dblp:conf/mm/ShenTHST16 fatcat:iscg57oryrhh5jnmvfnfji7xhe

Research Progress of Convolutional Neural Network and its Application in Object Detection [article]

Wei Zhang, Zuoxiang Zeng
2020 arXiv   pre-print
neural networks for object detection, pointing out the current deficiencies and future development direction.  ...  With the improvement of computer performance and the increase of data volume, the object detection based on convolutional neural network (CNN) has become the main algorithm for object detection.  ...  The model consists of a convolutional neural network (region proposal network, RPN) for extracting candidate regions and a convolutional neural network fast R-CNN for object detection.  ... 
arXiv:2007.13284v1 fatcat:wfv2zs6ozrd2vcsgoyq5sc42a4

Application of the transfer learning to the medical images texture classification task

M Privalov, M Stupina, V. Breskich, A. Zheltenkov, Y. Dreizis
2020 E3S Web of Conferences  
Presented work investigates different approaches to solving image classification task with neural networks, specifically, using pre-processing for feature extraction and end-to-end application of convolutional  ...  neural networks (CNN).  ...  Texture classification using pre-trained convolutional neural network and transfer learning.  ... 
doi:10.1051/e3sconf/202022401020 fatcat:kkztebtz7rbyvbamkgq5hjn6eu

Generating Binary Tags for Fast Medical Image Retrieval Based on Convolutional Nets and Radon Transform [article]

Xinran Liu, Hamid R. Tizhoosh, Jonathan Kofman
2016 arXiv   pre-print
Generally, different feature extraction methods are used to assign expressive and invariant features to each image such that the search for similar images comes down to feature classification and/or matching  ...  The present work introduces a new image retrieval method for medical applications that employs a convolutional neural network (CNN) with recently introduced Radon barcodes.  ...  METHOD Convolutional Neural Network Code (CNNC) A convolutional neural network (CNN) is a type of feedforward artificial neural network that is designed to use minimal image pre-processing and work on  ... 
arXiv:1604.04676v1 fatcat:lljgqct2szeoraifrzvfmjxfra

Untangling Local and Global Deformations in Deep Convolutional Networks for Image Classification and Sliding Window Detection [article]

George Papandreou and Iasonas Kokkinos and Pierre-André Savalle
2014 arXiv   pre-print
Deep Convolutional Neural Networks (DCNNs) commonly use generic 'max-pooling' (MP) layers to extract deformation-invariant features, but we argue in favor of a more refined treatment.  ...  For this we rely on a 'patchwork' data structure that efficiently lays out all image scales and positions as candidates to a DCNN.  ...  Acknowledgments We gratefully acknowledge the support of NVIDIA Corporation with the donation of GPUs used for this research.  ... 
arXiv:1412.0296v1 fatcat:vxx7eg6ilbcflftesxv26i2owy

Learning Transformation Invariant Representations with Weak Supervision

Benjamin Coors, Alexandru Condurache, Alfred Mertins, Andreas Geiger
2018 Proceedings of the 13th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications  
Deep convolutional neural networks are the current state-of-the-art solution to many computer vision tasks. However, their ability to handle large global and local image transformations is limited.  ...  Our loss acts as an effective regularizer which facilitates the learning of transformation invariant representations.  ...  Experimental Setup The rotated MNIST classification task (Larochelle et al., 2007) is the standard benchmark for evaluating transformation invariance in neural networks (Sohn and Lee, 2012; Cohen and  ... 
doi:10.5220/0006549000640072 dblp:conf/visapp/CoorsCMG18 fatcat:52zckkahn5esnnfsimqqbxpprm

CNN Architectures for Geometric Transformation-Invariant Feature Representation in Computer Vision: A Review

Alhassan Mumuni, Fuseini Mumuni
2021 SN Computer Science  
While convolutional neural networks (CNNs) have inherent representation power that provides a high degree of invariance to geometric image transformations, they are unable to satisfactorily handle nontrivial  ...  In addition, natural phenomena such as occlusion, deformation, and clutter can cause geometric appearance changes of the underlying objects, leading to geometric transformations of the resulting images  ...  In [118] Zhang et al. proposed a Multi-column Spatial Transformer Convolutional Neural Network (MC-STCNN) for traffic sign classification.  ... 
doi:10.1007/s42979-021-00735-0 fatcat:3zrkaan7dncoja4e32u7jgwo4m

Kernelized Deep Convolutional Neural Network for Describing Complex Images [article]

Zhen Liu
2015 arXiv   pre-print
With the impressive capability to capture visual content, deep convolutional neural networks (CNN) have demon- strated promising performance in various vision-based ap- plications, such as classification  ...  In this paper, to address this problem, we proposed a new kernelized deep convolutional neural network.  ...  convolutional neural network in detail.  ... 
arXiv:1509.04581v1 fatcat:yar3t2v2fbeipgzo3mrq22giq4

Vision Transformer with Convolutions Architecture Search [article]

Haichao Zhang, Kuangrong Hao, Witold Pedrycz, Lei Gao, Xuesong Tang, Bing Wei
2022 arXiv   pre-print
The high-performance backbone network searched by VTCAS introduces the desirable features of convolutional neural networks into the Transformer architecture while maintaining the benefits of the multi-head  ...  Therefore, here we take a step forward to study the structural characteristics of Transformer and convolution and propose an architecture search method-Vision Transformer with Convolutions Architecture  ...  Benchmark Datasets For image classification, the proposed image classification standard network searching by VTCAS is benchmarked on ImageNet-1K [1] , which contains 1.28M images and 50K validation images  ... 
arXiv:2203.10435v1 fatcat:cslntlodbrarhb66kshoqqrc2y

Equivariance-bridged SO(2)-Invariant Representation Learning using Graph Convolutional Network [article]

Sungwon Hwang, Hyungtae Lim, Hyun Myung
2021 arXiv   pre-print
Our method achieves the state-of-the-art image classification performance on rotated MNIST and CIFAR-10 images, where the models are trained with a non-augmented dataset only.  ...  Training a Convolutional Neural Network (CNN) to be robust against rotation has mostly been done with data augmentation.  ...  Conclusion We proposed a network that yields equivariant representation with SWN-GCN and invariant representation using GAP.  ... 
arXiv:2106.09996v2 fatcat:m7lv43z3urgspe2cvi7agta66y

Object Detection and Feature Base Learning with Sparse Convolutional Neural Networks [chapter]

Alexander R. T. Gepperth
2006 Lecture Notes in Computer Science  
A new convolutional neural network model termed sparse convolutional neural network (SCNN) is presented and its usefulness for real-time object detection in gray-valued, monocular video sequences is demonstrated  ...  SCNNs are trained on "raw" gray values and are intended to perform feature selection as a part of regular neural network training.  ...  In this article, a new convolutional neural network architecture termed sparse convolutional neural network (SCNN) is presented and its possibilities for object detection are explored.  ... 
doi:10.1007/11829898_20 fatcat:nezsgfjctna6bdtsveaj3gyr6e

Deep Learning Framework for Precipitation Prediction Using Cloud Images

Mirza Adnan Baig, Ghulam Ali Mallah, Noor Ahmed Shaikh
2022 Computers Materials & Continua  
Second step construct the decision model by using convolutional neural network (CNN) and third step presents the performance visualization by using confusion matrix, precision, recall and accuracy measures  ...  to consider for rainfall prediction.  ...  Acknowledgement: We shall remain thankful to the Department of Computer Science, Shah Abdul Latif University for providing a nice environment to conduct the research.  ... 
doi:10.32604/cmc.2022.026225 fatcat:ygpll7prw5bndjwsg4yaqehxv4

Some Improvements on Deep Convolutional Neural Network Based Image Classification [article]

Andrew G. Howard
2013 arXiv   pre-print
We investigate multiple techniques to improve upon the current state of the art deep convolutional neural network based image classification pipeline.  ...  The techiques include adding more image transformations to training data, adding more transformations to generate additional predictions at test time and using complementary models applied to higher resolution  ...  Since their introduction in the early 1990s [7] , convolutional neural networks have consistently been competitive with other techniques for image classification and recognition.  ... 
arXiv:1312.5402v1 fatcat:rkq27xddevbazikkosgjwjeevm

Convolutional Neural Networks Applied to Human Face Classification

Brian Cheung
2012 2012 11th International Conference on Machine Learning and Applications  
We trained a convolutional neural network to distinguish between images of human faces from computer generated avatars as part of the ICMLA 2012 Face Recognition Challenge.  ...  Convolutional neural network models have covered a broad scope of computer vision applications, achieving competitive performance with minimal domain knowledge.  ...  CONVOLUTIONAL NEURAL NETWORKS ConvNets have shown a great deal of promise in a wide variety of tasks from audio classification [15] to connectomics, a field which studies biological neural connectivity  ... 
doi:10.1109/icmla.2012.177 dblp:conf/icmla/Cheung12 fatcat:cwskodp2vzgptegc6iwcbunwdu
« Previous Showing results 1 — 15 out of 14,669 results