Rectified linear activation units are important components for state-of-the-art deep convolutional networks. In this paper, we propose a novel S-shaped rectifiedlinear activation unit (SReLU) to learn both convexand non-convex functions, imitating the multiple function forms given by the two fundamental laws, namely the Webner-Fechner law and the Stevens law, in psychophysics and neural sciences. Specifically, SReLU consists of three piecewise linear functions, which are formulated by four learnable parameters. The SReLU is learned jointly with the training of the whole deep network through back propagation. During the training phase, to initialize SReLU in different layers, we propose a “freezing” method to degenerate SReLU into a predefined leaky rectified linear unit in the initial several training epochs and then adaptively learn the good initial values. SReLU can be universally used in the existing deep networks with negligible additional parameters and computation cost. Experiments with two popular CNN architectures, Network in Network and GoogLeNet on scale-various benchmarks including CIFAR10, CIFAR100, MNIST and ImageNet demonstrate that SReLU achieves remarkable improvement compared to other activation functions.

Recently, deep learning methods that employ stacked denoising autoencoders (SDAs) have been successfully applied in domain adaptation. Remarkable performance in multi-domain sentiment analysis datasets has been reported, making deep learning a promising approach to domain adaptation problems. SDAs are distinguished by learning robust data representations for recovering the original features that have been artificially corrupted with noise. The idea has been further exploited to marginalize out the random corruptions by a state-of-the-art method called mSDA. In this paper, a deep learning method for domain adaptation called l 2,1 -norm stacked robust autoencoders ( l 2,1 -SRA) is proposed to learn useful representations for domain adaptation tasks. Each layer of l 2,1 -SRA contains two steps: a robust linear reconstruction step which is based on l 2,1 robust regression and a non-linear squashing transformation step. The experimental results demonstrate that the proposed method is very effective in multiple cross domain classification datasets which include Amazon review dataset, spam dataset from ECML/PKDD discovery challenge 2006 and 20 newsgroups dataset.

Obtaining a protein's 3D structure is crucial to the understanding of its functions and interactions with other proteins. It is critical to accelerate the protein crystallization process with improved accuracy for understanding cancer and designing drugs. Systematic high-throughput approaches in protein crystallization have been widely applied, generating a large number of protein crystallization-trial images. Therefore, an efficient and effective automatic analysis for these images is a top priority. In this paper, we present a novel system, CrystalNet, for automatically labeling outcomes of protein crystallization-trial images. CrystalNet is a deep convolutional neural network that automatically extracts features from X-ray protein crystallization images for classification. We show that (1) CrystalNet can provide real-time labels for crystallization images effectively, requiring approximately 2 seconds to provide labels for all 1536 images of crystallization microassay on each plate; (2) compared with the state-of-the-art classification systems in crystallization image analysis, our technique demonstrates an improvement of 8% in accuracy, and achieve 90.8% accuracy in classification. As a part of the high-throughput pipeline which generates millions of images a year, CrystalNet can lead to a substantial reduction of labor-intensive screening.

Photo privacy is a very important problem in the digital age where photos are commonly shared on social networking sites and mobile devices. The main challenge in photo privacy detection is how to generate discriminant features to accurately detect privacy at risk photos. Existing photo privacy detection works, which rely on low-level vision features, are non-informative to the users regarding what privacy information is leaked from their photos. In this paper, we propose a new framework called Privacy-CNH that utilizes hierarchical features which include both object and convolutional features in a deep learning model to detect privacy at risk photos. The generation of object features enables our model to better inform the users about the reason why a photo has privacy risk. The combination of convolutional and object features provide a richer model to understand photo privacy from different aspects, thus improving photo privacy detection accuracy. Experimental results demonstrate that the proposed model outperforms the state-of-the-art work and the standard convolutional neural network (CNN) with low-level features on photo privacy detection tasks.

In recent years, deep learning has spread beyond both academia and industry with many exciting real-world applications. The development of deep learning has presented obvious privacy issues. However, there has been lack of scientific study about privacy preservation in deep learning. In this paper, we concentrate on the auto-encoder, a fundamental component in deep learning, and propose the deep private auto-encoder (dPA). Our main idea is to enforce ε-differential privacy by perturbing the objective functions of the traditional deep auto-encoder, rather than its results. We apply the dPA to human behavior prediction in a health social network. Theoretical analysis and thorough experimental evaluations show that the dPA is highly effective and efficient, and it significantly outperforms existing solutions.

It is well established that in many scenarios there is no single solver that will provide optimal performance across a wide range of problem instances. Taking advantage of this observation, research into algorithm selection is designed to help identify the best approach for each problem at hand. This segregation is usually based on carefully constructed features, designed to quickly present the overall structure of the instance as a constant size numeric vector. Based on these features, a plethora of machine learning techniques can be utilized to predict the appropriate solver to execute, leading to significant improvements over relying solely on any one solver. However, being manually constructed, the creation of good features is an arduous task requiring a great deal of knowledge of the problem domain of interest. To alleviate this costly yet crucial step, this paper presents an automated methodology for producing an informative set of features utilizing a deep neural network. We show that the presented approach completely automates the algorithm selection pipeline and is able to achieve significantly better performance than a single best solver across multiple problem domains.

Style classification (e.g., architectural, music, fashion) attracts an increasing attention in both research and industrial fields. Most existing works focused on low-level visual features composition for style representation. However, little effort has been devoted to automatic mid-level or high-level style features learning by reorganizing low-level descriptors. Moreover, styles are usually spread out and not easy to differentiate from one to another. In this paper, we call these less representative images as weak style images. To address these issues, we propose a consensus style centralizing auto-encoder (CSCAE) to extract robust style features to facilitate weak style classification. CSCAE is the ensemble of several style centralizing auto-encoders (SCAEs) with consensus constraint. Each SCAE centralizes each feature of certain category in a progressive way. We apply our method in fashion style classification and manga style classification as two example applications. In addition, we collect a new dataset, Online Shopping, for fashion style classification evaluation, which will be publicly available for vision based fashion style research. Experiments demonstrate the effectiveness of SCAE and CSCAE on both public and newly collected datasets when compared with the most recent state-of-the-art works.

The goal of connectomics is to manifest the interconnections of neural system with the Electron Microscopy (EM) images. However, the formidable size of EM image data renders human annotation impractical, as it may take decades to fulfill the whole job. An alternative way to reconstruct the connectome can be attained with the computerized scheme that can automatically segment the neuronal structures. The segmentation of EM images is very challenging as the depicted structures can be very diverse.To address this difficult problem, a deep contextual network is proposed here by leveraging multi-level contextual information from the deep hierarchical structure to achieve better segmentation performance.To further improve the robustness against the vanishing gradients and strengthen the capability of the back-propagation of gradient flow, auxiliary classifiers are incorporated in the architecture of our deep neural network. It will be shown that our method can effectively parse the semantic meaning from the images with the underlying neural network and accurately delineate the structural boundaries with the reference of low-level contextual cues. Experimental results on the benchmark dataset of 2012 ISBI segmentation challenge of neuronal structures suggest that the proposed method can outperform the state-of-the-art methods by a large margin with respect to different evaluation measurements. Our method can potentially facilitate the automatic connectome analysis from EM images with less human intervention effort.

The number of mitoses per tissue area gives an important aggressiveness indication of the invasive breast carcinoma.However, automatic mitosis detection in histology images remains a challenging problem. Traditional methods either employ hand-crafted features to discriminate mitoses from other cells or construct a pixel-wise classifier to label every pixel in a sliding window way. While the former suffers from the large shape variation of mitoses and the existence of many mimics with similar appearance, the slow speed of the later prohibits its use in clinical practice.In order to overcome these shortcomings, we propose a fast and accurate method to detect mitosis by designing a novel deep cascaded convolutional neural network, which is composed of two components. First, by leveraging the fully convolutional neural network, we propose a coarse retrieval model to identify and locate the candidates of mitosis while preserving a high sensitivity.Based on these candidates, a fine discrimination model utilizing knowledge transferred from cross-domain is developed to further single out mitoses from hard mimics.Our approach outperformed other methods by a large margin in 2014 ICPR MITOS-ATYPIA challenge in terms of detection accuracy. When compared with the state-of-the-art methods on the 2012 ICPR MITOSIS data (a smaller and less challenging dataset), our method achieved comparable or better results with a roughly 60 times faster speed.

In this paper, we propose a novel model for learning graph representations, which generates a low-dimensional vector representation for each vertex by capturing the graph structural information. Different from other previous research efforts, we adopt a random surfing model to capture graph structural information directly, instead of using the sampling-based method for generating linear sequences proposed by Perozzi et al. (2014). The advantages of our approach will be illustrated from both theorical and empirical perspectives. We also give a new perspective for the matrix factorization method proposed by Levy and Goldberg (2014), in which the pointwise mutual information (PMI) matrix is considered as an analytical solution to the objective function of the skip-gram model with negative sampling proposed by Mikolov et al. (2013). Unlike their approach which involves the use of the SVD for finding the low-dimensitonal projections from the PMI matrix, however, the stacked denoising autoencoder is introduced in our model to extract complex features and model non-linearities. To demonstrate the effectiveness of our model, we conduct experiments on clustering and visualization tasks, employing the learned vertex representations as features. Empirical results on datasets of varying sizes show that our model outperforms other stat-of-the-art models in such tasks.