This paper presents a method for extracting distinctive invariant features from images that can be used to perform reliable matching between different views of an object or scene. The features are invariant to image scale and rotation, and are shown to provide robust matching across a substantial range of affine distortion, change in 3D viewpoint, addition of noise, and change in illumination. The features are highly distinctive, in the sense that a single feature can be correctly matched with high probability against a large database of features from many images. This paper also describes an approach to using these features for object recognition. The recognition proceeds by matching individual features to a database of features from known objects using a fast nearest-neighbor algorithm, followed by a Hough transform to identify clusters belonging to a single object, and finally performing verification through least-squares solution for consistent pose parameters. This approach to recognition can robustly identify objects among clutter and occlusion while achieving near real-time performance.

/pdf/distinctive-image-features-from-scale-invariant-keypoints-3waurjqzke.pdf

Distinctive Image Features from Scale-Invariant Keypoints

The increasing volume of data in modern business and science calls for more complex and sophisticated tools. Although advances in data mining technology have made extensive data collection much easier, it's still always evolving and there is a constant need for new techniques and tools that can help us transform this data into useful information and knowledge. Since the previous edition's publication, great advances have been made in the field of data mining. Not only does the third of edition of Data Mining: Concepts and Techniques continue the tradition of equipping you with an understanding and application of the theory and practice of discovering patterns hidden in large data sets, it also focuses on new, important topics in the field: data warehouses and data cube technology, mining stream, mining social networks, and mining spatial, multimedia and other complex data. Each chapter is a stand-alone guide to a critical topic, presenting proven algorithms and sound implementations ready to be used directly or with strategic modification against live data. This is the resource you need if you want to apply today's most powerful data mining techniques to meet real business challenges. * Presents dozens of algorithms and implementation examples, all in pseudo-code and suitable for use in real-world, large-scale data mining projects. * Addresses advanced topics such as mining object-relational databases, spatial databases, multimedia databases, time-series databases, text databases, the World Wide Web, and applications in several fields. *Provides a comprehensive, practical look at the concepts and techniques you need to get the most out of real business data

/pdf/data-mining-concepts-and-techniques-4dtvdfkvmi.pdf

Data Mining: Concepts and Techniques

Computer vision applications have come to rely increasingly on superpixels in recent years, but it is not always clear what constitutes a good superpixel algorithm. In an effort to understand the benefits and drawbacks of existing methods, we empirically compare five state-of-the-art superpixel algorithms for their ability to adhere to image boundaries, speed, memory efficiency, and their impact on segmentation performance. We then introduce a new superpixel algorithm, simple linear iterative clustering (SLIC), which adapts a k-means clustering approach to efficiently generate superpixels. Despite its simplicity, SLIC adheres to boundaries as well as or better than previous methods. At the same time, it is faster and more memory efficient, improves segmentation performance, and is straightforward to extend to supervoxel generation.

/pdf/slic-superpixels-compared-to-state-of-the-art-superpixel-5d8sq04am9.pdf

SLIC Superpixels Compared to State-of-the-Art Superpixel Methods

The k-means method is a widely used clustering technique that seeks to minimize the average squared distance between points in the same cluster. Although it offers no accuracy guarantees, its simplicity and speed are very appealing in practice. By augmenting k-means with a very simple, randomized seeding technique, we obtain an algorithm that is Θ(logk)-competitive with the optimal clustering. Preliminary experiments show that our augmentation improves both the speed and the accuracy of k-means, often quite dramatically.

/pdf/k-means-the-advantages-of-careful-seeding-4p8vluc5ni.pdf

k-means++: the advantages of careful seeding

http://library.utia.cas.cz/prace/20030125.pdf

Image registration methods: a survey

Symbolic pixel labeling for curvilinear feature detection

Traditional statistical methods deal with corroborating given hypotheses on a given body of data. However, generating the hypothesis itself is a matter of intuition and ingenuity. It is clearly impossible to test all hypotheses on a database with millions of records and hundreds of elds. There have been attempts to bridge this gap through data mining. Association generation is a method of creating such statistical hypotheses for binary data. For quantitative databases the situation is still not good. There are a number of known methods. One is a reduction to binary data by creating intervals and then generating associations. This method is computationally expensive. Another suggested method was by generating associations that are statistically interesting. This method also was tried only on small databases and is applicable only for binary relations, e.g., in certain ranges of eld X, eld Y lies signiicantly outside its average. We suggest a method that answers some of the problems with the current techniques. Our idea is based on using visualization techniques and image processing ideas to rank subsets of elds according to the relation between them in the database. This ranking suggests the hypotheses to be statistically investigated. Our method has the following advantages: 1. It is scalable. Our algorithm is mainly based on analyzing histograms of the data set, thus is more eecient. It is also naturally suitable for sampling. 2. It is generalizable in the size of the set of elds. No current method handles more than a binary relation. 3. It aaords comparability between elds over diierent base sets. This allows a uniform scale for diierent sets of elds in diierent databases. In this paper we present an algorithmic methodology and the results of its application to the census bureau data bases, cpsm93p and nhis93ac.

/pdf/analyzing-quantitative-databases-image-is-everything-20eb76z96h.pdf

Analyzing Quantitative Databases: Image is Everything

The nearest neighbor problem is that of preprocessing a set P of n data points in R d so that, given any query point q, the closest point in P to q can be determined eciently. In the chromatic nearest neighbor problem, each point of P is assigned a color, and the problem is to determine the color of the nearest point to the query point. More generally, given k 1, the problem is to determine the color occurring most frequently among the k nearest neighbors. The chromatic version of the nearest neighbor problem is used in many applications in pattern recognition and learning. In this paper we present a simple algorithm for solving the chromatic k nearest neighbor problem. We provide a query sensitive analysis, which shows that if the color classes form spatially well separated clusters (as often happens in practice), then queries can be answered quite eciently. We also allow the user to specify an error bound 0,

/pdf/chromatic-nearest-neighbor-searching-a-query-sensitive-4qx7l1buqy.pdf

Chromatic nearest neighbor searching: a query sensitive approach

We present an end-to-end learning method for chess, relying on deep neural networks. Without any a priori knowledge, in particular without any knowledge regarding the rules of chess, a deep neural network is trained using a combination of unsupervised pretraining and supervised training. The unsupervised training extracts high level features from a given position, and the supervised training learns to compare two chess positions and select the more favorable one. The training relies entirely on datasets of several million chess games, and no further domain specific knowledge is incorporated. 
The experiments show that the resulting neural network (referred to as DeepChess) is on a par with state-of-the-art chess playing programs, which have been developed through many years of manual feature selection and tuning. DeepChess is the first end-to-end machine learning-based method that results in a grandmaster-level chess playing performance.

DeepChess: End-to-End Deep Neural Network for Automatic Learning in Chess

Given a set P of n points in Rd, a fundamental problem in computational geometry is concerned with finding the smallest shape of some type that encloses all the points of P. Well-known instances of this problem include finding the smallest enclosing box, minimum volume ball, and minimum volume annulus. In this paper we consider the following variant: Given a set of n points in Rd, find the smallest shape in question that contains at least k points or a certain quantile of the data. This type of problem is known as a k-enclosing problem. We present a simple algorithmic framework for computing quantile approximations for the minimum strip, ellipsoid, and annulus containing a given quantile of the points. The algorithms run in O(n log n) time.

/pdf/quantile-approximation-for-robust-statistical-estimation-and-1ljr91ouwc.pdf

Nathan S. Netanyahu

Papers

Symbolic pixel labeling for curvilinear feature detection

Analyzing Quantitative Databases: Image is Everything

Chromatic nearest neighbor searching: a query sensitive approach

DeepChess: End-to-End Deep Neural Network for Automatic Learning in Chess

Quantile Approximation for Robust Statistical Estimation and k-Enclosing Problems