Memory-efficient Kernel PCA via Partial Matrix Sampling and Nonconvex Optimization: a Model-free Analysis of Local Minima

Kernel PCA is a widely used nonlinear dimension reduction technique in machine learning, but storing the kernel matrix is notoriously challenging when the sample size is large. Inspired by Yi et al. [2016], where the idea of partial matrix sampling followed by nonconvex optimization is proposed for matrix completion and robust PCA, we apply a similar approach to memory-efficient Kernel PCA. In theory, with no assumptions on the kernel matrix in terms of eigenvalues or eigenvectors, we established a model-free theory for the low-rank approximation based on any local minimum of the proposed objective function. As interesting byproducts, when the underlying positive semidefinite matrix is assumed to be low-rank and highly structured, corollaries of our main theorem improve the state-of-the-art results of Ge et al. [2016, 2017] for nonconvex matrix completion with no spurious local minima. Numerical experiments also show that our approach is competitive in terms of approximation accuracy compared to the well-known Nystr\"{o}m algorithm for Kernel PCA.

Predicting Aesthetic Score Distribution through Cumulative Jensen-Shannon Divergence

Aesthetic quality prediction is a challenging task in the computer vision community because of the complex interplay with semantic contents and photographic technologies. Recent studies on the powerful deep learning based aesthetic quality assessment usually use a binary high-low label or a numerical score to represent the aesthetic quality. However the scalar representation cannot describe well the underlying varieties of the human perception of aesthetics. In this work, we propose to predict the aesthetic score distribution (i.e., a score distribution vector of the ordinal basic human ratings) using Deep Convolutional Neural Network (DCNN). Conventional DCNNs which aim to minimize the difference between the predicted scalar numbers or vectors and the ground truth cannot be directly used for the ordinal basic rating distribution. Thus, a novel CNN based on the Cumulative distribution with Jensen-Shannon divergence (CJS-CNN) is presented to predict the aesthetic score distribution of human ratings, with a new reliability-sensitive learning method based on the kurtosis of the score distribution, which eliminates the requirement of the original full data of human ratings (without normalization). Experimental results on large scale aesthetic dataset demonstrate the effectiveness of our introduced CJS-CNN in this task.

3D Textured Model Encryption via 3D Lu Chaotic Mapping

In the coming Virtual/Augmented Reality (VR/AR) era, 3D contents will be popularized just as images and videos today. The security and privacy of these 3D contents should be taken into consideration. 3D contents contain surface models and solid models. The surface models include point clouds, meshes and textured models. Previous work mainly focus on encryption of solid models, point clouds and meshes. This work focuses on the most complicated 3D textured model. We propose a 3D Lu chaotic mapping based encryption method of 3D textured model. We encrypt the vertexes, the polygons and the textures of 3D models separately using the 3D Lu chaotic mapping. Then the encrypted vertices, edges and texture maps are composited together to form the final encrypted 3D textured model. The experimental results reveal that our method can encrypt and decrypt 3D textured models correctly. In addition, our method can resistant several attacks such as brute-force attack and statistic attack.

Single Reference Image based Scene Relighting via Material Guided Filtering

Image relighting is to change the illumination of an image to a target illumination effect without known the original scene geometry, material information and illumination condition. We propose a novel outdoor scene relighting method, which needs only a single reference image and is based on material constrained layer decomposition. Firstly, the material map is extracted from the input image. Then, the reference image is warped to the input image through patch match based image warping. Lastly, the input image is relit using material constrained layer decomposition. The experimental results reveal that our method can produce similar illumination effect as that of the reference image on the input image using only a single reference image.

Efficient Privacy Preserving Viola-Jones Type Object Detection via Random Base Image Representation

A cloud server spent a lot of time, energy and money to train a Viola-Jones type object detector with high accuracy. Clients can upload their photos to the cloud server to find objects. However, the client does not want the leakage of the content of his/her photos. In the meanwhile, the cloud server is also reluctant to leak any parameters of the trained object detectors. 10 years ago, Avidan & Butman introduced Blind Vision, which is a method for securely evaluating a Viola-Jones type object detector. Blind Vision uses standard cryptographic tools and is painfully slow to compute, taking a couple of hours to scan a single image. The purpose of this work is to explore an efficient method that can speed up the process. We propose the Random Base Image (RBI) Representation. The original image is divided into random base images. Only the base images are submitted randomly to the cloud server. Thus, the content of the image can not be leaked. In the meanwhile, a random vector and the secure Millionaire protocol are leveraged to protect the parameters of the trained object detector. The RBI makes the integral-image enable again for the great acceleration. The experimental results reveal that our method can retain the detection accuracy of that of the plain vision algorithm and is significantly faster than the traditional blind vision, with only a very low probability of the information leakage theoretically.

Convexified Modularity Maximization for Degree-corrected Stochastic Block Models

The stochastic block model (SBM) is a popular framework for studying community detection in networks. This model is limited by the assumption that all nodes in the same community are statistically equivalent and have equal expected degrees. The degree-corrected stochastic block model (DCSBM) is a natural extension of SBM that allows for degree heterogeneity within communities. This paper proposes a convexified modularity maximization approach for estimating the hidden communities under DCSBM. Our approach is based on a convex programming relaxation of the classical (generalized) modularity maximization formulation, followed by a novel doubly-weighted $ \ell_1 $-norm $ k $-median procedure. We establish non-asymptotic theoretical guarantees for both approximate clustering and perfect clustering. Our approximate clustering results are insensitive to the minimum degree, and hold even in sparse regime with bounded average degrees. In the special case of SBM, these theoretical results match the best-known performance guarantees of computationally feasible algorithms. Numerically, we provide an efficient implementation of our algorithm, which is applied to both synthetic and real-world networks. Experiment results show that our method enjoys competitive performance compared to the state of the art in the literature.

Optimal Rates of Convergence for Noisy Sparse Phase Retrieval via Thresholded Wirtinger Flow

This paper considers the noisy sparse phase retrieval problem: recovering a sparse signal $x \in \mathbb{R}^p$ from noisy quadratic measurements $y_j = (a_j' x )^2 + \epsilon_j$, $j=1, \ldots, m$, with independent sub-exponential noise $\epsilon_j$. The goals are to understand the effect of the sparsity of $x$ on the estimation precision and to construct a computationally feasible estimator to achieve the optimal rates. Inspired by the Wirtinger Flow [12] proposed for noiseless and non-sparse phase retrieval, a novel thresholded gradient descent algorithm is proposed and it is shown to adaptively achieve the minimax optimal rates of convergence over a wide range of sparsity levels when the $a_j$'s are independent standard Gaussian random vectors, provided that the sample size is sufficiently large compared to the sparsity of $x$.

Robust and computationally feasible community detection in the presence of arbitrary outlier nodes

Community detection, which aims to cluster $N$ nodes in a given graph into $r$ distinct groups based on the observed undirected edges, is an important problem in network data analysis. In this paper, the popular stochastic block model (SBM) is extended to the generalized stochastic block model (GSBM) that allows for adversarial outlier nodes, which are connected with the other nodes in the graph in an arbitrary way. Under this model, we introduce a procedure using convex optimization followed by $k$-means algorithm with $k=r$. Both theoretical and numerical properties of the method are analyzed. A theoretical guarantee is given for the procedure to accurately detect the communities with small misclassification rate under the setting where the number of clusters can grow with $N$. This theoretical result admits to the best-known result in the literature of computationally feasible community detection in SBM without outliers. Numerical results show that our method is both computationally fast and robust to different kinds of outliers, while some popular computationally fast community detection algorithms, such as spectral clustering applied to adjacency matrices or graph Laplacians, may fail to retrieve the major clusters due to a small portion of outliers. We apply a slight modification of our method to a political blogs data set, showing that our method is competent in practice and comparable to existing computationally feasible methods in the literature. To the best of the authors' knowledge, our result is the first in the literature in terms of clustering communities with fast growing numbers under the GSBM where a portion of arbitrary outlier nodes exist.

Compressed Sensing and Matrix Completion with Constant Proportion of Corruptions

We improve existing results in the field of compressed sensing and matrix completion when sampled data may be grossly corrupted. We introduce three new theorems. 1) In compressed sensing, we show that if the m \times n sensing matrix has independent Gaussian entries, then one can recover a sparse signal x exactly by tractable \ell1 minimimization even if a positive fraction of the measurements are arbitrarily corrupted, provided the number of nonzero entries in x is O(m/(log(n/m) + 1)). 2) In the very general sensing model introduced in "A probabilistic and RIPless theory of compressed sensing" by Candes and Plan, and assuming a positive fraction of corrupted measurements, exact recovery still holds if the signal now has O(m/(log^2 n)) nonzero entries. 3) Finally, we prove that one can recover an n \times n low-rank matrix from m corrupted sampled entries by tractable optimization provided the rank is on the order of O(m/(n log^2 n)); again, this holds when there is a positive fraction of corrupted samples.