Filters








5,350 Hits in 4.4 sec

Convolutional Neural Networks learn compact local image descriptors [article]

Christian Osendorfer, Justin Bayer, Patrick van der Smagt
2013 arXiv   pre-print
A standard deep convolutional neural network paired with a suitable loss function learns compact local image descriptors that perform comparably to state-of-the art approaches.  ...  We utilize DrLim to train a convolution neural network for learning lowdimensional mappings for low-level image patches.  ...  In this paper, f is a convolutional neural network (Jarrett et al., 2009) .  ... 
arXiv:1304.7948v2 fatcat:aqqk5tcbgzbuxlui4pct5lcgni

Convolutional Neural Networks Learn Compact Local Image Descriptors [chapter]

Christian Osendorfer, Justin Bayer, Sebastian Urban, Patrick van der Smagt
2013 Lecture Notes in Computer Science  
A standard deep convolutional neural network paired with a suitable loss function learns compact local image descriptors that perform comparably to state-of-the art approaches.  ...  We utilize DrLim to train a convolution neural network for learning lowdimensional mappings for low-level image patches.  ...  In this paper, f is a convolutional neural network (Jarrett et al., 2009) .  ... 
doi:10.1007/978-3-642-42051-1_77 fatcat:uhnexdak6ncvvanujj6ndrx3pm

Local Feature Descriptor Learning with Adaptive Siamese Network [article]

Chong Huang, Qiong Liu, Yan-Ying Chen, Kwang-Ting Cheng
2017 arXiv   pre-print
Specifically, the local feature is represented in a low dimensional space, so the neural network should have more compact structure.  ...  Although the recent progress in the deep neural network has led to the development of learnable local feature descriptors, there is no explicit answer for estimation of the necessary size of a neural network  ...  Specifically, the local feature is represented in a low dimensional space, so the neural network should have more compact structure.  ... 
arXiv:1706.05358v1 fatcat:kgipaoenarcypncstbthkyouxy

Locally-Transferred Fisher Vectors for Texture Classification

Yang Song, Fan Zhang, Qing Li, Heng Huang, Lauren J. ODonnell, Weidong Cai
2017 2017 IEEE International Conference on Computer Vision (ICCV)  
However, by truncating the CNN model at the last convolutional layer, the CNN-based FV descriptors would not incorporate the full capability of neural networks in feature learning.  ...  In particular, we design a locally-transferred Fisher vector (LFV) method, which involves a multi-layer neural network model containing locally connected layers to transform the input FV descriptors with  ...  from a convolutional neural network (CNN) model could produce more discriminative FV descriptors [7] .  ... 
doi:10.1109/iccv.2017.526 dblp:conf/iccv/SongZLHOC17 fatcat:aqofmp3s4vgkzno3mejvev5hn4

OREOS: Oriented Recognition of 3D Point Clouds in Outdoor Scenarios

Lukas Schaupp, Mathias Burki, Renaud Dube, Roland Siegwart, Cesar Cadena
2019 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)  
A Convolutional Neural Network is trained to extract compact descriptors from single 3D LiDAR scans.  ...  These can be used both to retrieve near-by place candidates from a map, and to estimate the yaw discrepancy needed for bootstrapping local registration methods.  ...  In a second stage, a Convolutional Neural Network is leveraged for extracting two compact descriptors v, and w.  ... 
doi:10.1109/iros40897.2019.8968094 dblp:conf/iros/SchauppBDSC19 fatcat:chckxkzkgfekvmefdb66emrai4

OREOS: Oriented Recognition of 3D Point Clouds in Outdoor Scenarios [article]

Lukas Schaupp, Mathias Bürki, Renaud Dubé, Roland Siegwart, Cesar Cadena
2019 arXiv   pre-print
A Convolutional Neural Network is trained to extract compact descriptors from single 3D LiDAR scans.  ...  These can be used both to retrieve near-by place candidates from a map, and to estimate the yaw discrepancy needed for bootstrapping local registration methods.  ...  In a second stage, a Convolutional Neural Network is leveraged for extracting two compact descriptors v, and w.  ... 
arXiv:1903.07918v1 fatcat:2scnozweabepziyh6cfl6vxph4

Supervised and Unsupervised End-to-End Deep Learning for Gene Ontology Classification of Neural In Situ Hybridization Images

Ido Cohen, Eli David, Nathan Netanyahu
2019 Entropy  
In contrast to traditional image processing methods, our method relies, instead, on deep convolutional denoising autoencoders (CDAE) for processing raw pixel inputs, and generating the desired compact  ...  In this paper, we describe a novel end-to-end deep learning-based method for generating compact representations of in situ hybridization (ISH) images, which are invariant-to-translation.  ...  [12] presents FuncISH, a learning method of functional representations of ISH images, using a histogram of local descriptors on several scales.  ... 
doi:10.3390/e21030221 pmid:33266936 fatcat:f7zlj5n7dbcgnmvgxuddc3kk6q

Feature Fusion for Robust Patch Matching With Compact Binary Descriptors [article]

Andrea Migliorati, Attilio Fiandrotti, Gianluca Francini, Skjalg Lepsoy, Riccardo Leonardi
2019 arXiv   pre-print
This work addresses the problem of learning compact yet discriminative patch descriptors within a deep learning framework.  ...  We propose a convolutional network framework for learning binary patch descriptors where pixel domain features are fused with features extracted from the transformed domain.  ...  Recent advances in deep learning showed that it is possible to train a neural network to automatically learn and compare local descriptors without necessarily resorting to handcrafted descriptors.  ... 
arXiv:1901.03547v1 fatcat:kbvmtmssuzbydiuta5paujqho4

Robust Line Segments Matching via Graph Convolution Networks [article]

QuanMeng Ma, Guang Jiang, DianZhi Lai
2020 arXiv   pre-print
In contrast to hand-crafted line matching algorithms, our approach learns local line segment descriptor and the matching simultaneously through end-to-end training.  ...  In this paper, we present a new method of using a graph convolution network to match line segments in a pair of images, and we design a graph-based strategy of matching line segments with relaxing to an  ...  First, to the best of our knowledge, we are the first to use the Convolutional Neural Networks and Graph Neural Networks to learn the local line descriptor and the matching in a unified end-to-end model  ... 
arXiv:2004.04993v2 fatcat:bzocewnyxbdejnkqy4bu7a5g7a

Learned BRIEF – transferring the knowledge from hand-crafted to learning-based descriptors

Nina Zizakic, Aleksandra Pizurica
2020 2020 IEEE 22nd International Workshop on Multimedia Signal Processing (MMSP)  
In this paper, we present a novel approach for designing local image descriptors that learn from data and from hand-crafted descriptors.  ...  We implement our learned BRIEF with a convolutional autoencoder architecture.  ...  Autoencoders Autoencoders [21] are unsupervised neural networks used for learning compact representations of data.  ... 
doi:10.1109/mmsp48831.2020.9287159 fatcat:i6ey7oinlnbxvebigalugxkqgy

DeepBrain: Functional Representation of Neural In-Situ Hybridization Images for Gene Ontology Classification Using Deep Convolutional Autoencoders [chapter]

Ido Cohen, Eli David, Nathan S. Netanyahu, Noa Liscovitch, Gal Chechik
2017 Lecture Notes in Computer Science  
This paper presents a novel deep learning-based method for learning a functional representation of mammalian neural images.  ...  The method uses a deep convolutional denoising autoencoder (CDAE) for generating an invariant, compact representation of in situ hybridization (ISH) images.  ...  In order to find a compact representation of these ISH images, we explored autoencoders (AE) and convolution neural networks (CNN), and found the convolutional autoencoder (CAE) to be the most appropriate  ... 
doi:10.1007/978-3-319-68612-7_33 fatcat:2og6qrydj5h5vbs3jg6cjgqq5y

Classification of Artistic Styles Using Binarized Features Derived from a Deep Neural Network [chapter]

Yaniv Bar, Noga Levy, Lior Wolf
2015 Lecture Notes in Computer Science  
The recent interest in deep neural networks has provided powerful visual features that achieve state-of-the-art results in various visual classification tasks.  ...  Combined with the PiCodes descriptors, these features show excellent classification results on a large scale collection of paintings.  ...  Convolutional neural networks (CNNs) are feed-forward networks that can be learned efficiently, and recent results indicate that the generic descriptors extracted from CNN are very powerful and provide  ... 
doi:10.1007/978-3-319-16178-5_5 fatcat:j7n3guyptjf6vmqrkcwwl6os3u

Comparing bags of features, conventional convolutional neural network and AlexNet for fruit recognition

Nik Noor Akmal Abdul Hamid, Rabiatul Adawiya Razali, Zaidah Ibrahim
2019 Indonesian Journal of Electrical Engineering and Computer Science  
This paper presents a comparative study between Bag of Features (BoF), Conventional Convolutional Neural Network (CNN) and Alexnet for fruit recognition.  ...  Due to the outstanding performance of deep learning like CNN and its pre-trained models like AlexNet in image recognition, this paper investigates the accuracy of conventional CNN, and Alexnet in recognizing  ...  As they are represented as histograms of local descriptors, BoF gives an extremely compact description of images.  ... 
doi:10.11591/ijeecs.v14.i1.pp333-339 fatcat:ccbss7uawfeabntsfusovwdmgm

Group Invariant Deep Representations for Image Instance Retrieval [article]

Olivier Morère, Antoine Veillard, Jie Lin, Julie Petta, Vijay Chandrasekhar, Tomaso Poggio
2016 arXiv   pre-print
Due to their success in large scale image classification, representations extracted from Convolutional Neural Networks (CNN) are quickly gaining ground on Fisher Vectors (FVs) as state-of-the-art global  ...  Our method implements a recently proposed mathematical theory for invariance in a sensory cortex modeled as a feedforward neural network.  ...  Index Terms-Invariance theory, group invariance, deep learning, convolutional neural networks, image instance retrieval, global image descriptors, compact descriptors. I.  ... 
arXiv:1601.02093v2 fatcat:x4k5sntopzg27mvw2sssy36ydm

Compact Deep Color Features for Remote Sensing Scene Classification

Rao Muhammad Anwer, Fahad Shahbaz Khan, Jorma Laaksonen
2021 Neural Processing Letters  
Most recent aerial scene classification approaches are based on Convolutional Neural Networks (CNNs).  ...  However, the importance of color within the deep learning framework is yet to be investigated for aerial scene classification.  ...  The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material.  ... 
doi:10.1007/s11063-021-10463-4 fatcat:getvq2myhvayhbgzdzpknzenkq
« Previous Showing results 1 — 15 out of 5,350 results