We propose a deep convolutional neural network architecture codenamed Inception that achieves the new state of the art for classification and detection in the ImageNet Large-Scale Visual Recognition Challenge 2014 (ILSVRC14). The main hallmark of this architecture is the improved utilization of the computing resources inside the network. By a carefully crafted design, we increased the depth and width of the network while keeping the computational budget constant. To optimize quality, the architectural decisions were based on the Hebbian principle and the intuition of multi-scale processing. One particular incarnation used in our submission for ILSVRC14 is called GoogLeNet, a 22 layers deep network, the quality of which is assessed in the context of classification and detection.

/pdf/going-deeper-with-convolutions-1yobw2o2ds.pdf

Going deeper with convolutions

There is, I think, something ethereal about i —the square root of minus one. I remember first hearing about it at school. It seemed an odd beast at that time—an intruder hovering on the edge of reality.

Usually familiarity dulls this sense of the bizarre, but in the case of i it was the reverse: over the years the sense of its surreal nature intensified. It seemed that it was impossible to write mathematics that described the real world in …

I and i

Machine Learning is the study of methods for programming computers to learn. Computers are applied to a wide range of tasks, and for most of these it is relatively easy for programmers to design and implement the necessary software. However, there are many tasks for which this is difficult or impossible. These can be divided into four general categories. First, there are problems for which there exist no human experts. For example, in modern automated manufacturing facilities, there is a need to predict machine failures before they occur by analyzing sensor readings. Because the machines are new, there are no human experts who can be interviewed by a programmer to provide the knowledge necessary to build a computer system. A machine learning system can study recorded data and subsequent machine failures and learn prediction rules. Second, there are problems where human experts exist, but where they are unable to explain their expertise. This is the case in many perceptual tasks, such as speech recognition, hand-writing recognition, and natural language understanding. Virtually all humans exhibit expert-level abilities on these tasks, but none of them can describe the detailed steps that they follow as they perform them. Fortunately, humans can provide machines with examples of the inputs and correct outputs for these tasks, so machine learning algorithms can learn to map the inputs to the outputs. Third, there are problems where phenomena are changing rapidly. In finance, for example, people would like to predict the future behavior of the stock market, of consumer purchases, or of exchange rates. These behaviors change frequently, so that even if a programmer could construct a good predictive computer program, it would need to be rewritten frequently. A learning program can relieve the programmer of this burden by constantly modifying and tuning a set of learned prediction rules. Fourth, there are applications that need to be customized for each computer user separately. Consider, for example, a program to filter unwanted electronic mail messages. Different users will need different filters. It is unreasonable to expect each user to program his or her own rules, and it is infeasible to provide every user with a software engineer to keep the rules up-to-date. A machine learning system can learn which mail messages the user rejects and maintain the filtering rules automatically. Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis. Statistics focuses on understanding the phenomena that have generated the data, often with the goal of testing different hypotheses about those phenomena. Data mining seeks to find patterns in the data that are understandable by people. Psychological studies of human learning aspire to understand the mechanisms underlying the various learning behaviors exhibited by people (concept learning, skill acquisition, strategy change, etc.).

Machine learning

Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Pattern Recognition and Machine Learning

计量经济分析 = Econometric analysis

In the work presented here we propose a novel bicoherence-based method for the classification of aerial radar targets in automatic target recognition (ATR) systems. The possibility of classifying aerial targets using the micro-Doppler contributions caused by a jet engine or the rotor of a helicopter is studied. The method is based on classification features computed in the form of bicoherence estimates, as well as cepstral coefficients extracted from the micro-Doppler contribution contained in radar returns. The performance of the classification method developed is compared with the performance of common methods using high-resolution radar range profiles (HRRPs). Correct classification probability rates are computed for three different types of aerial targets. The benefits achieved by using bicoherence-based classification features are demonstrated and discussed.

Classification of aircraft using micro-Doppler bicoherence-based features

Image Filtering Based on Discrete Cosine Transform

Local Approximation Techniques in Signal and Image Processing (SPIE Press Monograph Vol. PM157)

We introduce a new transform domain (least mean square) LMS algorithm with variable step. The existing approaches use different time-variable step-sizes for each filter tap. The step-sizes are time-variable due to the power estimates of each transform coefficient. In our new approach, for each step-size we define a local component that is given by the power normalization, and a global component that is the same for each filter coefficient. We show that if the global component is also made time-variable, depending on the output error, the speed of convergence can be significantly improved.

A transform domain LMS adaptive filter with variable step-size

In this paper, a novel single image super-resolution (SISR) algorithm is proposed. It is based on the BM3D (Block-Matching and 3D filtering) paradigm, where both sparsity and nonlocal patch self-similarity priors are utilized. The algorithm is derived from a variational formulation of the problem and has a structure typical for iterative back-projection super-resolution methods. They are characterized by updating high-resolution image which is calculated using the previous estimate and upsampled low-resolution error. The developed method is thoroughly compared with the state-of-the-art SISR both for noiseless and noisy data, demonstrating superior performance objectively and subjectively.

Karen Egiazarian

Papers

Classification of aircraft using micro-Doppler bicoherence-based features

Image Filtering Based on Discrete Cosine Transform

Local Approximation Techniques in Signal and Image Processing (SPIE Press Monograph Vol. PM157)

A transform domain LMS adaptive filter with variable step-size

Single image super-resolution via BM3D sparse coding