Deep learning for visual understanding: A review
NeurocomputingPublished 27 November 2015
Yanming Guo, Yu Liu, Ard Oerlemans, Songyang Lao, Song Wu, Michael S. Lew
Citations2,451
SJR quartileQ1
SJR score1.47
SNIP1.94
Generate an AI Snapshot to get a quick, structured summary of this paper.
Study Snapshot
ObjectiveStudy objective
MethodsResearch methodology
PopulationPopulation studied
Sample sizeSample sizes
OutcomesStudy outcomes here
ResultsStudy results comes here
LimitationsResearch study limitations comes here
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
The state-of-the-art in deep learning algorithms in computer vision is reviewed by highlighting the contributions and challenges from over 210 recent research papers, and the future trends and challenges in designing and training deep neural networks are summarized.
Abstract
Deep learning algorithms are a subset of the machine learning algorithms, which aim at discovering multiple levels of distributed representations. Recently, numerous deep learning algorithms have been proposed to solve traditional artificial intelligence problems. This work aims to review the state-of-the-art in deep learning algorithms in computer vision by highlighting the contributions and challenges from over 210 recent research papers. It first gives an overview of various deep learning approaches and their recent developments, and then briefly describes their applications in diverse vision tasks, such as image classification, object detection, image retrieval, semantic segmentation and human pose estimation. Finally, the paper summarizes the future trends and challenges in designing and training deep neural networks.
Keywords
Computer Science
Communications of the ACMImageNet classification with deep convolutional neural networks
75,709 Citations2017Alex Krizhevsky, Ilya Sutskever +1 more
A large, deep convolutional neural network was trained to classify the 1.2 million high-resolution images in the ImageNet LSVRC-2010 contest into the 1000 different classes and employed a recently developed regularization method called "dropout" that proved to be very effective.
arXiv (Cornell University)Very Deep Convolutional Networks for Large-Scale Image Recognition
75,540 Citations2014Karen Simonyan, Andrew Zisserman
This work investigates the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting using an architecture with very small convolution filters, which shows that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 weight layers.
2009 IEEE Conference on Computer Vision and Pattern RecognitionImageNet: A large-scale hierarchical image database
61,841 Citations2009Jia Deng, Wei Dong +4 more
A new database called “ImageNet” is introduced, a large-scale ontology of images built upon the backbone of the WordNet structure, much larger in scale and diversity and much more accurate than the current image datasets.
Proceedings of the IEEEGradient-based learning applied to document recognition
58,219 Citations1998Yann LeCun, Léon Bottou +2 more
This paper reviews various methods applied to handwritten character recognition and compares them on a standard handwritten digit recognition task, and Convolutional neural networks are shown to outperform all other techniques.
Going deeper with convolutions
46,825 Citations2015Christian Szegedy, Wei Liu +7 more
A deep convolutional neural network architecture codenamed Inception is proposed that achieves the new state of the art for classification and detection in the ImageNet Large-Scale Visual Recognition Challenge 2014 (ILSVRC14).
International Journal of Computer VisionImageNet Large Scale Visual Recognition Challenge
40,207 Citations2015Olga Russakovsky, Jia Deng +10 more
The creation of this benchmark dataset and the advances in object recognition that have been possible as a result are described, and the state-of-the-art computer vision accuracy with human accuracy is compared.
Fully convolutional networks for semantic segmentation
36,888 Citations2015Jonathan Long, Evan Shelhamer +1 more
The key insight is to build “fully convolutional” networks that take input of arbitrary size and produce correspondingly-sized output with efficient inference and learning.
Dropout: a simple way to prevent neural networks from overfitting
34,279 Citations2014Nitish Srivastava, Geoffrey E. Hinton +3 more
It is shown that dropout improves the performance of neural networks on supervised learning tasks in vision, speech recognition, document classification and computational biology, obtaining state-of-the-art results on many benchmark data sets.
Histograms of Oriented Gradients for Human Detection
31,896 Citations2005Navneet Dalal, Bill Triggs
It is shown experimentally that grids of histograms of oriented gradient (HOG) descriptors significantly outperform existing feature sets for human detection, and the influence of each stage of the computation on performance is studied.
Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation
31,752 Citations2014Ross Girshick, Jeff Donahue +2 more
This paper proposes a simple and scalable detection algorithm that improves mean average precision (mAP) by more than 30% relative to the previous best result on VOC 2012 -- achieving a mAP of 53.3%.
NatureLearning representations by back-propagating errors
30,885 Citations1986David E. Rumelhart, Geoffrey E. Hinton +1 more
Back-propagation repeatedly adjusts the weights of the connections in the network so as to minimize a measure of the difference between the actual output vector of the net and the desired output vector, which helps to represent important features of the task domain.
Fast R-CNN
27,899 Citations2015Ross Girshick
arXiv (Cornell University)Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
24,381 Citations2015Sergey Ioffe, Christian Szegedy
Applied to a state-of-the-art image classification model, Batch Normalization achieves the same accuracy with 14 times fewer training steps, and beats the original model by a significant margin.
ScienceReducing the Dimensionality of Data with Neural Networks
20,953 Citations2006Geoffrey E. Hinton, Ruslan Salakhutdinov
Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification
18,777 Citations2015Kaiming He, Xiangyu Zhang +2 more
This work proposes a Parametric Rectified Linear Unit (PReLU) that generalizes the traditional rectified unit and derives a robust initialization method that particularly considers the rectifier nonlinearities.
arXiv (Cornell University)Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
18,238 Citations2015Shaoqing Ren, Kaiming He +2 more
arXiv (Cornell University)Distributed Representations of Words and Phrases and their Compositionality
18,086 Citations2013Tomáš Mikolov, Ilya Sutskever +3 more
This paper presents a simple method for finding phrases in text, and shows that learning good vector representations for millions of phrases is possible and describes a simple alternative to the hierarchical softmax called negative sampling.
Neural NetworksDeep learning in neural networks: An overview
18,033 Citations2014Jürgen Schmidhuber
This historical survey compactly summarizes relevant work, much of it from the previous millennium, review deep supervised learning, unsupervised learning, reinforcement learning & evolutionary computation, and indirect search for short programs encoding deep and large networks.
Neural ComputationA Fast Learning Algorithm for Deep Belief Nets
16,404 Citations2006Geoffrey E. Hinton, Simon Osindero +1 more
A fast, greedy algorithm is derived that can learn deep, directed belief networks one layer at a time, provided the top two layers form an undirected associative memory.
Lecture notes in computer scienceVisualizing and Understanding Convolutional Networks
15,444 Citations2014Matthew D. Zeiler, Rob Fergus
A novel visualization technique is introduced that gives insight into the function of intermediate feature layers and the operation of the classifier in large Convolutional Network models, used in a diagnostic role to find model architectures that outperform Krizhevsky et al on the ImageNet classification benchmark.
International Conference on Machine LearningRectified Linear Units Improve Restricted Boltzmann Machines
13,244 Citations2010Vinod Nair, Geoffrey E. Hinton
Restricted Boltzmann machines were developed using binary stochastic hidden units that learn features that are better for object recognition on the NORB dataset and face verification on the Labeled Faces in the Wild dataset.
IEEE Transactions on Pattern Analysis and Machine IntelligenceRepresentation Learning: A Review and New Perspectives
13,002 Citations2013Yoshua Bengio, Aaron Courville +1 more
Recent work in the area of unsupervised feature learning and deep learning is reviewed, covering advances in probabilistic models, autoencoders, manifold learning, and deep networks.
A training algorithm for optimal margin classifiers
11,594 Citations1992Bernhard E. Boser, Isabelle Guyon +1 more
A training algorithm that maximizes the margin between the training patterns and the decision boundary is presented, applicable to a wide variety of the classification functions, including Perceptrons, polynomials, and Radial Basis Functions.
IEEE Transactions on Pattern Analysis and Machine IntelligenceObject Detection with Discriminatively Trained Part-Based Models
9,976 Citations2009Pedro F. Felzenszwalb, Ross Girshick +2 more
An object detection system based on mixtures of multiscale deformable part models that is able to represent highly variable object classes and achieves state-of-the-art results in the PASCAL object detection challenges is described.
Beyond Bags of Features: Spatial Pyramid Matching for Recognizing Natural Scene Categories
7,929 Citations2006Svetlana Lazebnik, C. Schmid +1 more
This paper presents a method for recognizing scene categories based on approximate global geometric correspondence that exceeds the state of the art on the Caltech-101 database and achieves high accuracy on a large database of fifteen natural scene categories.
Extracting and composing robust features with denoising autoencoders
7,343 Citations2008Pascal Vincent, Hugo Larochelle +2 more
This work introduces and motivate a new training principle for unsupervised learning of a representation based on the idea of making the learned representations robust to partial corruption of the input pattern.
arXiv (Cornell University)Improving neural networks by preventing co-adaptation of feature detectors
6,631 Citations2012Geoffrey E. Hinton, Nitish Srivastava +3 more
International Journal of Computer VisionSelective Search for Object Recognition
6,118 Citations2013Jasper Uijlings, Koen E. A. van de Sande +2 more
This paper introduces selective search which combines the strength of both an exhaustive search and segmentation, and shows that its selective search enables the use of the powerful Bag-of-Words model for recognition.
Stacked Denoising Autoencoders: Learning Useful Representations in a Deep Network with a Local Denoising Criterion
5,014 Citations2010Pascal Vincent, Hugo Larochelle +3 more
CNN Features Off-the-Shelf: An Astounding Baseline for Recognition
4,341 Citations2014Ali Sharif Razavian, Hossein Azizpour +2 more
A series of experiments conducted for different recognition tasks using the publicly available code and model of the OverFeat network which was trained to perform object classification on ILSVRC13 suggest that features obtained from deep learning with convolutional nets should be the primary candidate in most visual recognition tasks.
Visual categorization with bags of keypoints
4,147 Citations2004G. Csurka
Journal of Machine Learning ResearchStacked Denoising Autoencoders: Learning Useful Representations in a Deep Network with a Local Denoising Criterion
4,118 Citations2010VincentPascal, LarochelleHugo +3 more
This work clearly establishes the value of using a denoising criterion as a tractable unsupervised objective to guide the learning of useful higher level representations.
Learning Deconvolution Network for Semantic Segmentation
4,036 Citations2015Hyeonwoo Noh, Seunghoon Hong +1 more
A novel semantic segmentation algorithm by learning a deep deconvolution network on top of the convolutional layers adopted from VGG 16-layer net, which demonstrates outstanding performance in PASCAL VOC 2012 dataset.
Multi-column deep neural networks for image classification
3,764 Citations2012Dan Cireşan, Ueli Meier +1 more
On the very competitive MNIST handwriting benchmark, this method is the first to achieve near-human performance and improves the state-of-the-art on a plethora of common image classification benchmarks.
Vision ResearchSparse coding with an overcomplete basis set: A strategy employed by V1?
3,720 Citations1997Bruno A. Olshausen, David J. Field
These deviations from linearity provide a potential explanation for the weak forms of non-linearity observed in the response properties of cortical simple cells, and they further make predictions about the expected interactions among units in response to naturalistic stimuli.
arXiv (Cornell University)Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs
3,622 Citations2014Liang-Chieh Chen, George Papandreou +3 more
arXiv (Cornell University)How transferable are features in deep neural networks?
3,453 Citations2014Jason Yosinski, Jeff Clune +2 more
Deep neural networks are easily fooled: High confidence predictions for unrecognizable images
3,260 Citations2015Anh‐Tu Nguyen, Jason Yosinski +1 more
This work takes convolutional neural networks trained to perform well on either the ImageNet or MNIST datasets and finds images with evolutionary algorithms or gradient ascent that DNNs label with high confidence as belonging to each dataset class, and produces fooling images, which are then used to raise questions about the generality of DNN computer vision.
DeepPose: Human Pose Estimation via Deep Neural Networks
3,254 Citations2014Alexander Toshev, Christian Szegedy
The pose estimation is formulated as a DNN-based regression problem towards body joints and a cascade of such DNN regres- sors which results in high precision pose estimates.
You Only Look Once: Unified, Real-Time Object Detection
3,253 Citations2016Joseph Redmon, Santosh Divvala +2 more
Compared to state-of-the-art detection systems, YOLO makes more localization errors but is less likely to predict false positives on background, and outperforms other detection methods, including DPM and R-CNN, when generalizing from natural images to other domains like artwork.
Learning and Transferring Mid-level Image Representations Using Convolutional Neural Networks
3,158 Citations2014Maxime Oquab, Léon Bottou +2 more
This work designs a method to reuse layers trained on the ImageNet dataset to compute mid-level image representation for images in the PASCAL VOC dataset, and shows that despite differences in image statistics and tasks in the two datasets, the transferred representation leads to significantly improved results for object and action classification.
Lecture notes in computer scienceSpatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition
3,113 Citations2014Kaiming He, Xiangyu Zhang +2 more
This work equips the networks with another pooling strategy, “spatial pyramid pooling”, to eliminate the above requirement, and develops a new network structure, called SPP-net, which can generate a fixed-length representation regardless of image size/scale.
Holistically-Nested Edge Detection
3,067 Citations2015Saining Xie, Zhuowen Tu
Locality-constrained Linear Coding for image classification
3,061 Citations2010Jinjun Wang, Shuicheng Yan +4 more
This paper presents a simple but effective coding scheme called Locality-constrained Linear Coding (LLC) in place of the VQ coding in traditional SPM, using the locality constraints to project each descriptor into its local-coordinate system, and the projected coordinates are integrated by max pooling to generate the final representation.
2009 IEEE Conference on Computer Vision and Pattern RecognitionLinear spatial pyramid matching using sparse coding for image classification
2,869 Citations2009Shuicheng Yan, Kai Yu +2 more
An extension of the SPM method is developed, by generalizing vector quantization to sparse coding followed by multi-scale spatial max pooling, and a linear SPM kernel based on SIFT sparse codes is proposed, leading to state-of-the-art performance on several benchmarks by using a single type of descriptors.
Lecture notes in computer scienceA Practical Guide to Training Restricted Boltzmann Machines
2,802 Citations2012Geoffrey E. Hinton
This guide is an attempt to share expertise at training restricted Boltzmann machines with other machine learning researchers.
Lecture notes in computer scienceEdge Boxes: Locating Object Proposals from Edges
2,766 Citations2014C. Lawrence Zitnick, Piotr Dollár
A novel method for generating object bounding box proposals using edges is proposed, showing results that are significantly more accurate than the current state-of-the-art while being faster to compute.
Lecture notes in computer scienceImproving the Fisher Kernel for Large-Scale Image Classification
2,609 Citations2010Florent Perronnin, Jorge Sánchez +1 more
In an evaluation involving hundreds of thousands of training images, it is shown that classifiers learned on Flickr groups perform surprisingly well and that they can complement classifier learned on more carefully annotated datasets.
Communications on Pure and Applied MathematicsFor most large underdetermined systems of linear equations the minimal 𝓁<sub>1</sub>‐norm solution is also the sparsest solution
2,541 Citations2006David L. Donoho
The techniques include the use of random proportional embeddings and almost‐spherical sections in Banach space theory, and deviation bounds for the eigenvalues of random Wishart matrices.
Convolutional deep belief networks for scalable unsupervised learning of hierarchical representations
2,446 Citations2009Honglak Lee, Roger Grosse +2 more
The convolutional deep belief network is presented, a hierarchical generative model which scales to realistic image sizes and is translation-invariant and supports efficient bottom-up and top-down probabilistic inference.
The MIT Press eBooksEfficient sparse coding algorithms
2,393 Citations2007Honglak Lee, Alexis Battle +2 more
These algorithms are applied to natural images and it is demonstrated that the inferred sparse codes exhibit end-stopping and non-classical receptive field surround suppression and, therefore, may provide a partial explanation for these two phenomena in V1 neurons.
Conditional Random Fields as Recurrent Neural Networks
2,391 Citations2015Shuai Zheng, Sadeep Jayasumana +6 more
A new form of convolutional neural network that combines the strengths of Convolutional Neural Networks (CNNs) and Conditional Random Fields (CRFs)-based probabilistic graphical modelling is introduced, and top results are obtained on the challenging Pascal VOC 2012 segmentation benchmark.
Journal of Machine Learning ResearchOnline Learning for Matrix Factorization and Sparse Coding
2,338 Citations2010Julien Mairal, Francis Bach +2 more
International Journal of Computer VisionPictorial Structures for Object Recognition
2,199 Citations2004Pedro F. Felzenszwalb, Daniel P. Huttenlocher
A computationally efficient framework for part-based modeling and recognition of objects, motivated by the pictorial structure models introduced by Fischler and Elschlager, that allows for qualitative descriptions of visual appearance and is suitable for generic recognition problems.
Why Does Unsupervised Pre-training Help Deep Learning?
2,115 Citations2010Dumitru Erhan, Aaron Courville +2 more
Online dictionary learning for sparse coding
2,111 Citations2009Julien Mairal, Francis Bach +2 more
A new online optimization algorithm for dictionary learning is proposed, based on stochastic approximations, which scales up gracefully to large datasets with millions of training samples, and leads to faster performance and better dictionaries than classical batch algorithms for both small and large datasets.
Lecture notes in computer scienceStacked Convolutional Auto-Encoders for Hierarchical Feature Extraction
1,992 Citations2011Jonathan Masci, Ueli Meier +2 more
A novel convolutional auto-encoder (CAE) for unsupervised feature learning that initializing a CNN with filters of a trained CAE stack yields superior performance on a digit and an object recognition benchmark.
The Annals of Applied StatisticsPathwise coordinate optimization
1,942 Citations2007Jerome H. Friedman, Trevor Hastie +2 more
It is shown that coordinate descent is very competitive with the well-known LARS procedure in large lasso problems, can deliver a path of solutions efficiently, and can be applied to many other convex statistical problems such as the garotte and elastic net.
International review of cytologyRegularization of Neural Networks using DropConnect
1,916 Citations2013Li Wan, Matthew D. Zeiler +3 more
This work introduces DropConnect, a generalization of Dropout, for regularizing large fully-connected layers within neural networks, and derives a bound on the generalization performance of both Dropout and DropConnect.
Deconvolutional networks
1,688 Citations2010Matthew D. Zeiler, Dilip Krishnan +2 more
This work presents a learning framework where features that capture these mid-level cues spontaneously emerge from image data, based on the convolutional decomposition of images under a spar-sity constraint and is totally unsupervised.
An HOG-LBP human detector with partial occlusion handling
1,688 Citations2009Xiaoyu Wang, Tony Xiao Han +1 more
By combining Histograms of Oriented Gradients (HOG) and Local Binary Pattern (LBP) as the feature set, this work proposes a novel human detection approach capable of handling partial occlusion and achieves the best human detection performance on the INRIA dataset.
Building high-level features using large scale unsupervised learning
1,650 Citations2013Quoc V. Le
Contrary to what appears to be a widely-held intuition, the experimental results reveal that it is possible to train a face detector without having to label images as containing a face or not.
Hypercolumns for object segmentation and fine-grained localization
1,578 Citations2015Bharath Hariharan, Pablo Arbeláez +2 more
Using hypercolumns as pixel descriptors, this work defines the hypercolumn at a pixel as the vector of activations of all CNN units above that pixel, and shows results on three fine-grained localization tasks: simultaneous detection and segmentation, and keypoint localization.
Self-taught learning
1,533 Citations2007Rajat Raina, Alexis Battle +3 more
An approach to self-taught learning that uses sparse coding to construct higher-level features using the unlabeled data to form a succinct input representation and significantly improve classification performance.
Lecture notes in computer scienceEvaluation of Pooling Operations in Convolutional Architectures for Object Recognition
1,518 Citations2010Dominik Scherer, Andreas Müller +1 more
The aim is to gain insight into different functions by directly comparing them on a fixed architecture for several common object recognition tasks, and empirical results show that a maximum pooling operation significantly outperforms subsampling operations.
Convolutional neural networks at constrained time cost
1,505 Citations2015Kaiming He, Jian Sun
This paper investigates the accuracy of CNNs under constrained time cost, and presents an architecture that achieves very competitive accuracy in the ImageNet dataset, yet is 20% faster than “AlexNet” [14] (16.0% top-5 error, 10-view test).
Deep Convolutional Network Cascade for Facial Point Detection
1,436 Citations2013Yi Sun, Xiaogang Wang +1 more
The proposed approach outperforms state-of-the-art methods in both detection accuracy and reliability and can avoid local minimum caused by ambiguity and data corruption in difficult image samples due to occlusions, large pose variations, and extreme lightings.
Efficient object localization using Convolutional Networks
1,359 Citations2015Jonathan Tompson, Ross Goroshin +3 more
A novel architecture which includes an efficient `position refinement' model that is trained to estimate the joint offset location within a small region of the image to achieve improved accuracy in human joint location estimation is introduced.
arXiv (Cornell University)OverFeat: Integrated Recognition, Localization and Detection using Convolutional Networks
1,332 Citations2013Pierre Sermanet, David Eigen +4 more
Exploiting Generative Models in Discriminative Classifiers
1,285 Citations1998Tommi Jaakkola, David Haussler
A natural way of achieving this combination by deriving kernel functions for use in discriminative methods such as support vector machines from generative probability models is developed.
Learning Fast Approximations of Sparse Coding
1,260 Citations2010Karol Gregor, Yann LeCun
Two versions of a very fast algorithm that produces approximate estimates of the sparse code that can be used to compute good visual features, or to initialize exact iterative algorithms are proposed.
Adaptive deconvolutional networks for mid and high level feature learning
1,249 Citations2011Matthew D. Zeiler, Graham W. Taylor +1 more
A hierarchical model that learns image decompositions via alternating layers of convolutional sparse coding and max pooling, relying on a novel inference scheme that ensures each layer reconstructs the input, rather than just the output of the layer directly beneath, as is common with existing hierarchical approaches.
IEEE Transactions on Pattern Analysis and Machine IntelligenceMeasuring the Objectness of Image Windows
1,231 Citations2012Bogdan Alexe, Thomas Deselaers +1 more
A generic objectness measure, quantifying how likely it is for an image window to contain an object of any class, and uses objectness as a complementary score in addition to the class-specific model, which leads to fewer false positives.
Lecture notes in computer scienceNeural Codes for Image Retrieval
1,219 Citations2014Artem Babenko, Anton Slesarev +2 more
It is established that neural codes perform competitively even when the convolutional neural network has been trained for an unrelated classification task (e.g. Image-Net), and the improvement in the retrieval performance of neural codes, when the network is retrained on a dataset of images that are similar to images encountered at test time.
Deep Neural Networks for Object Detection
1,206 Citations2013Christian Szegedy, Alexander Toshev +1 more
This paper presents a simple and yet powerful formulation of object detection as a regression problem to object bounding box masks, and defines a multi-scale inference procedure which is able to produce high-resolution object detections at a low cost by a few network applications.
International Joint Conference on Artificial IntelligenceFlexible, high performance convolutional neural networks for image classification
1,203 Citations2011Dan Cireşan, Ueli Meier +3 more
Scalable Object Detection Using Deep Neural Networks
1,157 Citations2014Dumitru Erhan, Christian Szegedy +2 more
This work proposes a saliency-inspired neural network model for detection, which predicts a set of class-agnostic bounding boxes along with a single score for each box, corresponding to its likelihood of containing any object of interest.
A Theoretical Analysis of Feature Pooling in Visual Recognition
1,148 Citations2010Y-Lan Boureau, Jean Ponce +1 more
It is shown that the reasons underlying the performance of various pooling methods are obscured by several confounding factors, such as the link between the sample cardinality in a spatial pool and the resolution at which low-level features have been extracted.
Contractive Auto-Encoders: Explicit Invariance During Feature Extraction
1,136 Citations2011Salah Rifai, Pascal Vincent +3 more
It is found empirically that this penalty helps to carve a representation that better captures the local directions of variation dictated by the data, corresponding to a lower-dimensional non-linear manifold, while being more invariant to the vast majority of directions orthogonal to the manifold.
Lecture notes in computer scienceSimultaneous Detection and Segmentation
1,122 Citations2014Bharath Hariharan, Pablo Arbeláez +2 more
This work builds on recent work that uses convolutional neural networks to classify category-independent region proposals (R-CNN), introducing a novel architecture tailored for SDS, and uses category-specific, top-down figure-ground predictions to refine the bottom-up proposals.
IEEE Computational Intelligence MagazineDeep Machine Learning - A New Frontier in Artificial Intelligence Research [Research Frontier]
1,110 Citations2010Itamar Arel, Derek Rose +1 more
An overview of the mainstream deep learning approaches and research directions proposed over the past decade is provided and some perspective into how it may evolve is presented.
The MIT Press eBooksEfficient Learning of Sparse Representations with an Energy-Based Model
1,079 Citations2007Marc’Aurelio Ranzato, Christopher S. Poultney +2 more
A novel unsupervised method for learning sparse, overcomplete features using a linear encoder, and a linear decoder preceded by a sparsifying non-linearity that turns a code vector into a quasi-binary sparse code vector.
BING: Binarized Normed Gradients for Objectness Estimation at 300fps
1,073 Citations2014Ming‐Ming Cheng, Ziming Zhang +2 more
Weakly-and Semi-Supervised Learning of a Deep Convolutional Network for Semantic Image Segmentation
1,068 Citations2015George Papandreou, Liang-Chieh Chen +2 more
Expectation-Maximization (EM) methods for semantic image segmentation model training under weakly supervised and semi-supervised settings are developed and extensive experimental evaluation shows that the proposed techniques can learn models delivering competitive results on the challenging PASCAL VOC 2012 image segmentsation benchmark, while requiring significantly less annotation effort.
BoxSup: Exploiting Bounding Boxes to Supervise Convolutional Networks for Semantic Segmentation
1,054 Citations2015Jifeng Dai, Kaiming He +1 more
This paper proposes a method that achieves competitive accuracy but only requires easily obtained bounding box annotations, and yields state-of-the-art results on PASCAL VOC 2012 and PASCal-CONTEXT.
Lecture notes in computer scienceMulti-scale Orderless Pooling of Deep Convolutional Activation Features
1,048 Citations2014Yunchao Gong, Liwei Wang +2 more
A simple but effective scheme called multi-scale orderless pooling (MOP-CNN), which extracts CNN activations for local patches at multiple scale levels, performs orderless VLAD pooling of these activations at each level separately, and concatenates the result.
Neural NetworksMulti-column deep neural network for traffic sign classification
1,039 Citations2012Dan Cireşan, Ueli Meier +2 more
This work uses a fast, fully parameterizable GPU implementation of a Deep Neural Network (DNN) that does not require careful design of pre-wired feature extractors, which are rather learned in a supervised way.
arXiv (Cornell University)Joint Training of a Convolutional Network and a Graphical Model for Human Pose Estimation
976 Citations2014Jonathan Tompson, Arjun Jain +2 more
This paper proposes a new hybrid architecture that consists of a deep Convolu-tional Network and a Markov Random Field and shows how this architecture is successfully applied to the challenging problem of articulated human pose estimation in monocular images.
HD-CNN: Hierarchical Deep Convolutional Neural Networks for Large Scale Visual Recognition
922 Citations2015Zhicheng Yan, Hao Zhang +5 more
This paper introduces hierarchical deepCNNs (HD-CNNs) by embedding deep CNNs into a two-level category hierarchy and achieves state-of-the-art results on both CIFAR100 and large-scale ImageNet 1000-class benchmark datasets.
Is object localization for free? - Weakly-supervised learning with convolutional neural networks
916 Citations2015Maxime Oquab, Léon Bottou +2 more
A weakly supervised convolutional neural network is described for object classification that relies only on image-level labels, yet can learn from cluttered scenes containing multiple objects.
PLoS Computational BiologyDeep Neural Networks Rival the Representation of Primate IT Cortex for Core Visual Object Recognition
901 Citations2014Charles F. Cadieu, Ha Hong +6 more
These evaluations show that, unlike previous bio-inspired models, the latest DNNs rival the representational performance of IT cortex on this visual object recognition task and propose an extension of “kernel analysis” that measures the generalization accuracy as a function of representational complexity.
International Conference on Machine LearningOn optimization methods for deep learning
888 Citations2011Jiquan Ngiam, Adam Coates +4 more
It is shown that more sophisticated off-the-shelf optimization methods such as Limited memory BFGS (L-BFGS) and Conjugate gradient (CG) with line search can significantly simplify and speed up the process of pretraining deep algorithms.
Clustered Pose and Nonlinear Appearance Models for Human Pose Estimation
884 Citations2010P. Sam Johnson, Mark Everingham
A new annotated database of challenging consumer images is introduced, an order of magnitude larger than currently available datasets, and over 50% relative improvement in pose estimation accuracy over a state-of-the-art method is demonstrated.
Sparse deep belief net model for visual area V2
881 Citations2007Honglak Lee, Chaitanya Ekanadham +1 more
An unsupervised learning model is presented that faithfully mimics certain properties of visual area V2 and the encoding of these more complex "corner" features matches well with the results from the Ito & Komatsu's study of biological V2 responses, suggesting that this sparse variant of deep belief networks holds promise for modeling more higher-order features.
Efficient Piecewise Training of Deep Structured Models for Semantic Segmentation
847 Citations2016Guosheng Lin, Chunhua Shen +2 more
This work shows how to improve semantic segmentation through the use of contextual information, specifically, ' patch-patch' context between image regions, and 'patch-background' context, and formulate Conditional Random Fields with CNN-based pairwise potential functions to capture semantic correlations between neighboring patches.
Computers & Mathematics with ApplicationsParallel optimization: Theory, algorithms, and applications
830 Citations1998
Deep Learning for Content-Based Image Retrieval
810 Citations2014Ji Wan, D. Wang +5 more
This paper investigates a framework of deep learning with application to CBIR tasks with an extensive set of empirical studies by examining a state-of-the-art deep learning method (Convolutional Neural Networks) for CBIr tasks under varied settings.
Object Detection via a Multi-region and Semantic Segmentation-Aware CNN Model
803 Citations2015Spyros Gidaris, Nikos Komodakis
An object detection system that relies on a multi-region deep convolutional neural network that also encodes semantic segmentation-aware features that aims at capturing a diverse set of discriminative appearance factors and exhibits localization sensitivity that is essential for accurate object localization.
…