From the Publisher:
This is the first comprehensive treatment of feed-forward neural networks from the perspective of statistical pattern recognition. After introducing the basic concepts, the book examines techniques for modelling probability density functions and the properties and merits of the multi-layer perceptron and radial basis function network models. Also covered are various forms of error functions, principal algorithms for error function minimalization, learning and generalization in neural networks, and Bayesian techniques and their applications. Designed as a text, with over 100 exercises, this fully up-to-date work will benefit anyone involved in the fields of neural computation and pattern recognition.

/pdf/neural-networks-for-pattern-recognition-34puigiau1.pdf

Neural networks for pattern recognition

The core component of most modern trackers is a discriminative classifier, tasked with distinguishing between the target and the surrounding environment. To cope with natural image changes, this classifier is typically trained with translated and scaled sample patches. Such sets of samples are riddled with redundancies—any overlapping pixels are constrained to be the same. Based on this simple observation, we propose an analytic model for datasets of thousands of translated patches. By showing that the resulting data matrix is circulant, we can diagonalize it with the discrete Fourier transform, reducing both storage and computation by several orders of magnitude. Interestingly, for linear regression our formulation is equivalent to a correlation filter, used by some of the fastest competitive trackers. For kernel regression, however, we derive a new kernelized correlation filter (KCF), that unlike other kernel algorithms has the exact same complexity as its linear counterpart. Building on it, we also propose a fast multi-channel extension of linear correlation filters, via a linear kernel, which we call dual correlation filter (DCF). Both KCF and DCF outperform top-ranking trackers such as Struck or TLD on a 50 videos benchmark, despite running at hundreds of frames-per-second, and being implemented in a few lines of code (Algorithm 1). To encourage further developments, our tracking framework was made open-source.

High-Speed Tracking with Kernelized Correlation Filters

We present a neural network-based upright frontal face detection system. A retinally connected neural network examines small windows of an image and decides whether each window contains a face. The system arbitrates between multiple networks to improve performance over a single network. We present a straightforward procedure for aligning positive face examples for training. To collect negative examples, we use a bootstrap algorithm, which adds false detections into the training set as training progresses. This eliminates the difficult task of manually selecting nonface training examples, which must be chosen to span the entire space of nonface images. Simple heuristics, such as using the fact that faces rarely overlap in images, can further improve the accuracy. Comparisons with several other state-of-the-art face detection systems are presented, showing that our system has comparable performance in terms of detection and false-positive rates.

Neural network-based face detection

A major new professional reference work on fingerprint security systems and technology from leading international researchers in the field Handbook provides authoritative and comprehensive coverage of all major topics, concepts, and methods for fingerprint security systems This unique reference work is an absolutely essential resource for all biometric security professionals, researchers, and systems administrators

/pdf/handbook-of-fingerprint-recognition-16mh887j24.pdf

Handbook of Fingerprint Recognition

http://romisatriawahono.net/lecture/rm/survey/machine%20learning/Chandrashekar%20-%20Feature%20Selection%20Methods%20-%202014.pdf

A survey on feature selection methods

Real-time optical inspection techniques are discussed and implemented on laboratory systems. Specifically, we show how Fourier transform and Hough transform data can be used in an industrial inspection application. We also discuss how a 1-D synthetic discriminant function is used for label inspection. Real-time optical Hough transform inspection data using a 2-D spatial light modulator is presented and discussed. A color liquid crystal television (LCTV) is used as the input transducer to provide real-time data. New methods to process color image data on color LCTVs are detailed. A new high speed acoustooptic architecture to generate Hough transform slices is also detailed. A specific product inspection case study is discussed.

Industrial use of a real-time optical inspection system.

Classification and rejection tests were performed on the 10-class MSTAR database. To compare our performance, we first summarize relevant prior work. Shift-invariant magnitude Fourier transform (FT) features were used in the feature space trajectory (FST) classifier to classify the 10-class MSTAR data with variants and to reject confuser images and clutter chips. No prior work has addressed this. Implication of various SAR preprocessing on performance is addressed. In confuser rejection, 2 standard confusers (used in prior work) and 2 new confusers are addressed. We are the first to extend confuser rejection tests to 8 target classification with variants and 2 confuser rejection. Finally, clutter rejection while classifying the 10 targets is addressed.

Classification and rejection of MSTAR data

Segmentation of agricultural products on a conveyor belt is a first step for product inspection. We consider a new algorithm to achieve segmentation (locating each individual item and placing it in a separate file). We use X-ray images which provide useful internal information on the state of the product (good or bad). The product items considered vary in size, shape, gray-scale properties, and lie at random orientations. Many items touch and thus new techniques are necessary to detect each item at any orientation, estimate the number of touching items and their centers, and segment overlapping touching items to provide separate image files for touching input items. Test results are presented for X-ray film and linescan images of various agricultural products including one extensive database.

Modified binary watershed transform for segmentation of agricultural products

Two optical implementations of production systems are advanced. The production systems operate on a knowledge base where facts and rules are encoded as formulas in propositional calculus. The first implementation is a binary neural network. An analog neural network is used to include reasoning with uncertainties. The second implementation uses a new optical symbolic substitution correlator. This implementation is useful when a set of similar situations has to be handled in parallel on one processor.

David P. Casasent

Papers

Industrial use of a real-time optical inspection system.

Classification and rejection of MSTAR data

Modified binary watershed transform for segmentation of agricultural products

Optical production systems using neural networks and symbolic substitution

Direct and implicit optical matrix–vector algorithms: addendum