Posted Content•
Neural Architecture Search with Reinforcement Learning
TL;DR: This paper uses a recurrent network to generate the model descriptions of neural networks and trains this RNN with reinforcement learning to maximize the expected accuracy of the generated architectures on a validation set.
Abstract: Neural networks are powerful and flexible models that work well for many difficult learning tasks in image, speech and natural language understanding. Despite their success, neural networks are still hard to design. In this paper, we use a recurrent network to generate the model descriptions of neural networks and train this RNN with reinforcement learning to maximize the expected accuracy of the generated architectures on a validation set. On the CIFAR-10 dataset, our method, starting from scratch, can design a novel network architecture that rivals the best human-invented architecture in terms of test set accuracy. Our CIFAR-10 model achieves a test error rate of 3.65, which is 0.09 percent better and 1.05x faster than the previous state-of-the-art model that used a similar architectural scheme. On the Penn Treebank dataset, our model can compose a novel recurrent cell that outperforms the widely-used LSTM cell, and other state-of-the-art baselines. Our cell achieves a test set perplexity of 62.4 on the Penn Treebank, which is 3.6 perplexity better than the previous state-of-the-art model. The cell can also be transferred to the character language modeling task on PTB and achieves a state-of-the-art perplexity of 1.214.
Citations
More filters
TL;DR: In this article, an ameliorated deep dense convolutional neural network (BX-Net) was presented to accurately recognize casting defects in X-ray images and effectively extract highly discriminative features of different categories.
Abstract: Recognizing defects in X-ray images plays an important role in the detection of internal defects in titanium alloy castings. However, the existing manual defects recognition methods have common drawbacks such as unstable artificial recognition, misrecognition, huge workload, and low efficiency of recognition. To make up for the shortcomings, an ameliorated deep dense convolutional neural network (BX-Net) was presented to accurately recognize casting defects in X-ray images and effectively extract highly discriminative features of different categories. DenseNet121 was used as the backbone of BX-Net and the feature extracted by DenseNet121 was fully shared by the two inputs of a bilinear pooling layer. Transfer learning was applied to reduce the demand for data and hyperparameters’ tuning. The backbone of BX-Net was firstly trained on the ImageNet and then all layers of BX-Net was fine-tuned on nine-hundred X-ray images of TiAl aero casting components. Other six deep convolutional neural networks (DenseNet121, EfficientNetB4, EfficientNetB7, ResNet50, VGG16 and Xception) were also trained to be compared with the presented BX-Net. Two Support Vector Machines were trained on the LBP features data set and HOG features data set of nine-hundred X-ray images of TiAl aero casting components respectively Experiments comparing BX-Net with other six deep learning models and two machine learning models on one hundred X-ray images (test set) of TiAl aero casting components were carried out. The comparison results show that BX-Net has the least parameters except for the Densenet121. The recall and accuracy of BX-Net were 99% and 99% respectively. In addition, the comparison results also show that BX-Net was the only model that learned discriminative feature representation of the casting X-ray image data set. The BX-Net proposed in this paper is expected to overcome the shortcomings of manual defects recognition.
15 citations
Patent•
19 Jul 201815 citations
Proceedings Article•
03 Dec 2018TL;DR: In this paper, the authors cast the problem of neural network structure learning as a problem of Bayesian structure learning, and instead of directly learning the discriminative structure, it learns a generative graph, constructs its stochastic inverse, and then constructs a discriminant graph.
Abstract: We introduce a principled approach for unsupervised structure learning of deep neural networks. We propose a new interpretation for depth and inter-layer connectivity where conditional independencies in the input distribution are encoded hierarchically in the network structure. Thus, the depth of the network is determined inherently. The proposed method casts the problem of neural network structure learning as a problem of Bayesian network structure learning. Then, instead of directly learning the discriminative structure, it learns a generative graph, constructs its stochastic inverse, and then constructs a discriminative graph. We prove that conditional-dependency relations among the latent variables in the generative graph are preserved in the class-conditional discriminative graph. We demonstrate on image classification benchmarks that the deepest layers (convolutional and dense) of common networks can be replaced by significantly smaller learned structures, while maintaining classification accuracy—state-of-the-art on tested benchmarks. Our structure learning algorithm requires a small computational cost and runs efficiently on a standard desktop CPU.
15 citations
10 Jun 2019
TL;DR: This paper proposes a novel method based on genetic programming to simultaneously find the optimal activation functions and optimization techniques and analyzes the activation function found and confirms the usefulness of the proposed method.
Abstract: Recently, deep learning is one of the most popular techniques in artificial intelligence. However, to construct a deep learning model, various components must be set up, including activation functions, optimization methods, a configuration of model structure called hyperparameters. As they affect the performance of deep learning, researchers are working hard to find optimal hyperparameters when solving problems with deep learning. Activation function and optimization technique play a crucial role in the forward and backward processes of model learning, but they are set up in a heuristic way. The previous studies have been conducted to optimize either activation function or optimization technique, while the relationship between them is neglected to search them at the same time. In this paper, we propose a novel method based on genetic programming to simultaneously find the optimal activation functions and optimization techniques. In genetic programming, each individual is composed of two chromosomes, one for the activation function and the other for the optimization technique. To calculate the fitness of one individual, we construct a neural network with the activation function and optimization technique that the individual represents. The deep learning model found through our method has 82.59% and 53.04% of accuracies for the CIFAR-10 and CIFAR-100 datasets, which outperforms the conventional methods. Moreover, we analyze the activation function found and confirm the usefulness of the proposed method.
15 citations
Cites background from "Neural Architecture Search with Rei..."
...It is not surprising, however, to use it to automatically find the formula for them as it solves many problems with deep learning [10-14]....
[...]
Posted Content•
TL;DR: This paper provides an introduction to the most relevant deep learning models and concepts, point out possible pitfalls by analyzing special characteristics of SAR data, review the state-of-the-art of deep learning applied to SAR in depth, summarize available benchmarks, and recommend some important future research directions.
Abstract: Deep learning in remote sensing has become an international hype, but it is mostly limited to the evaluation of optical data. Although deep learning has been introduced in Synthetic Aperture Radar (SAR) data processing, despite successful first attempts, its huge potential remains locked. In this paper, we provide an introduction to the most relevant deep learning models and concepts, point out possible pitfalls by analyzing special characteristics of SAR data, review the state-of-the-art of deep learning applied to SAR in depth, summarize available benchmarks, and recommend some important future research directions. With this effort, we hope to stimulate more research in this interesting yet under-exploited research field and to pave the way for use of deep learning in big SAR data processing workflows.
15 citations
Cites methods from "Neural Architecture Search with Rei..."
...The first major attempt in this field was by Zoph et al. [44], who used DeepRL to find the optimum CNN for image classification....
[...]
...Recently DeepRL received particular attention and popularity due to the success of Google Deep Mind’s AlphaGo [43], which defeated the Go board game world champion....
[...]
...DeepRL uses different types of neural networks to create these functions [41][42]....
[...]
...3) Deep Reinforcement Learning (DeepRL): Reinforcement Learning (RL) tries to mimic the human learning behavior, i.e., taking actions and then adjusting them for the future according to feedback from the environment....
[...]
...Sources of the images: VGG [15], ResNet [16], U-Net [17], LSTM [18], RNN [19], VAE [20], GAN [21], CGNN [22], RGNN [23], and DeepRL [24]....
[...]
References
More filters
27 Jun 2016
TL;DR: In this article, the authors proposed a residual learning framework to ease the training of networks that are substantially deeper than those used previously, which won the 1st place on the ILSVRC 2015 classification task.
Abstract: Deeper neural networks are more difficult to train. We present a residual learning framework to ease the training of networks that are substantially deeper than those used previously. We explicitly reformulate the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. We provide comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth. On the ImageNet dataset we evaluate residual nets with a depth of up to 152 layers—8× deeper than VGG nets [40] but still having lower complexity. An ensemble of these residual nets achieves 3.57% error on the ImageNet test set. This result won the 1st place on the ILSVRC 2015 classification task. We also present analysis on CIFAR-10 with 100 and 1000 layers. The depth of representations is of central importance for many visual recognition tasks. Solely due to our extremely deep representations, we obtain a 28% relative improvement on the COCO object detection dataset. Deep residual nets are foundations of our submissions to ILSVRC & COCO 2015 competitions1, where we also won the 1st places on the tasks of ImageNet detection, ImageNet localization, COCO detection, and COCO segmentation.
123,388 citations
Proceedings Article•
01 Jan 2015TL;DR: This work introduces Adam, an algorithm for first-order gradient-based optimization of stochastic objective functions, based on adaptive estimates of lower-order moments, and provides a regret bound on the convergence rate that is comparable to the best known results under the online convex optimization framework.
Abstract: We introduce Adam, an algorithm for first-order gradient-based optimization of stochastic objective functions, based on adaptive estimates of lower-order moments. The method is straightforward to implement, is computationally efficient, has little memory requirements, is invariant to diagonal rescaling of the gradients, and is well suited for problems that are large in terms of data and/or parameters. The method is also appropriate for non-stationary objectives and problems with very noisy and/or sparse gradients. The hyper-parameters have intuitive interpretations and typically require little tuning. Some connections to related algorithms, on which Adam was inspired, are discussed. We also analyze the theoretical convergence properties of the algorithm and provide a regret bound on the convergence rate that is comparable to the best known results under the online convex optimization framework. Empirical results demonstrate that Adam works well in practice and compares favorably to other stochastic optimization methods. Finally, we discuss AdaMax, a variant of Adam based on the infinity norm.
111,197 citations
Proceedings Article•
04 Sep 2014TL;DR: 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.
Abstract: In this work we investigate the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting. Our main contribution is a thorough evaluation of networks of increasing depth using an architecture with very small (3x3) 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. These findings were the basis of our ImageNet Challenge 2014 submission, where our team secured the first and the second places in the localisation and classification tracks respectively. We also show that our representations generalise well to other datasets, where they achieve state-of-the-art results. We have made our two best-performing ConvNet models publicly available to facilitate further research on the use of deep visual representations in computer vision.
55,235 citations
"Neural Architecture Search with Rei..." refers methods in this paper
...Along with this success is a paradigm shift from feature designing to architecture designing, i.e., from SIFT (Lowe, 1999), and HOG (Dalal & Triggs, 2005), to AlexNet (Krizhevsky et al., 2012), VGGNet (Simonyan & Zisserman, 2014), GoogleNet (Szegedy et al., 2015), and ResNet (He et al., 2016a)....
[...]
01 Jan 1998
TL;DR: In this article, a graph transformer network (GTN) is proposed for handwritten character recognition, which can be used to synthesize a complex decision surface that can classify high-dimensional patterns, such as handwritten characters.
Abstract: Multilayer neural networks trained with the back-propagation algorithm constitute the best example of a successful gradient based learning technique. Given an appropriate network architecture, gradient-based learning algorithms can be used to synthesize a complex decision surface that can classify high-dimensional patterns, such as handwritten characters, with minimal preprocessing. This paper reviews various methods applied to handwritten character recognition and compares them on a standard handwritten digit recognition task. Convolutional neural networks, which are specifically designed to deal with the variability of 2D shapes, are shown to outperform all other techniques. Real-life document recognition systems are composed of multiple modules including field extraction, segmentation recognition, and language modeling. A new learning paradigm, called graph transformer networks (GTN), allows such multimodule systems to be trained globally using gradient-based methods so as to minimize an overall performance measure. Two systems for online handwriting recognition are described. Experiments demonstrate the advantage of global training, and the flexibility of graph transformer networks. A graph transformer network for reading a bank cheque is also described. It uses convolutional neural network character recognizers combined with global training techniques to provide record accuracy on business and personal cheques. It is deployed commercially and reads several million cheques per day.
42,067 citations
20 Jun 2005
TL;DR: 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.
Abstract: We study the question of feature sets for robust visual object recognition; adopting linear SVM based human detection as a test case. After reviewing existing edge and gradient based descriptors, we show experimentally that grids of histograms of oriented gradient (HOG) descriptors significantly outperform existing feature sets for human detection. We study the influence of each stage of the computation on performance, concluding that fine-scale gradients, fine orientation binning, relatively coarse spatial binning, and high-quality local contrast normalization in overlapping descriptor blocks are all important for good results. The new approach gives near-perfect separation on the original MIT pedestrian database, so we introduce a more challenging dataset containing over 1800 annotated human images with a large range of pose variations and backgrounds.
31,952 citations
"Neural Architecture Search with Rei..." refers methods in this paper
...Along with this success is a paradigm shift from feature designing to architecture designing, i.e., from SIFT (Lowe, 1999), and HOG (Dalal & Triggs, 2005), to AlexNet (Krizhevsky et al., 2012), VGGNet (Simonyan & Zisserman, 2014), GoogleNet (Szegedy et al., 2015), and ResNet (He et al., 2016a)....
[...]