Use of log polar space for foveation and feature recognition

The finding and maintaining of high accuracy foveation points for several types of recognised feature in log polar space such as a line, circular or elliptical arc is considered. Log polar space is preferred over cartesian space as it provides a high resolution and a wide viewing angle; feature invariance in the fovea simplifies foveation; it allows multi-resolution analysis; and rotation and scale are linear translations in log polar space.

An investigation into the use of physical modelling for the prediction of various feature types visible from different view points

This paper describes a general purpose flexible technique which uses physical modelling techniques for determining the features of a 3D object that are visible from any predefined view. Physical modelling techniques are used to determine which of many different types of features are visible from a complete set of viewpoints. The power of this technique lies in its ability to detect and parameterise object features, regardless of object complexity. Raytracing is used to simulate the physical process by which object features are visible so that surface properties (eg specularity, transparency) as well as object boundaries can be used in the recognition process. Using this technique occluding and non-occluding edge based features are extracted using image processing techniques and then parameterised. Features caused by specularity are also extracted and qualitative descriptions for these are defined.

Comparative evaluation of pedestrian detection methods for mobile bus surveillance

We present a comparative evaluation of the state-of-art algorithms for detecting pedestrians in low frame rate and low resolution footage acquired by mobile sensors. Four approaches are compared: a) The Histogram of Oriented Gradient (HoG) approach [1]; b) A new histogram feature that is formed by the weighted sum of both the gradient magnitude and the filter responses from a set of elongated Gaussian filters [2] corresponding to the quantised orientation, called Histogram of Oriented Gradient Banks (HoGB) approach; c) The codebook based HoG feature with branch-and-bound (efficient subwindow search) algorithm [3] and; d) The codebook based HoGB approach. Results show that the HoG based detector achieves the highest performance in terms of the true positive detection, the HoGB approach has the lowest false positives whilst maintaining a comparable true positive rate to the HoG, and the codebook approaches allow computationally efficient detection.

Generating comprehensible summaries of rushes sequences based on robust feature matching

This paper describes our first attempt at tackling a pilot task in Trecvid: video summarization of rushes data [3]. Our method is based on the tight clustering produced via SIFT matching. In this first attempt, we try to examine how our approach performs without complex implementation in terms of concept detection and excerpt assembly (i.e, no picture-in-picture, split screen and special transitions). Although we do not perform very well in terms of concept inclusion, we rank very well in terms of the summary being easy to understand and relevancy of included segments.

Feature reduction to speed up malware classification

In statistical classification work, one method of speeding up the process is to use only a small percentage of the total parameter set available. In this paper, we apply this technique both to the classification of malware and the identification of malware from a set combined with cleanware. In order to demonstrate the usefulness of our method, we use the same sets of malware and cleanware as in an earlier paper. Using the statistical technique Information Gain (IG), we reduce the set of features used in the experiment from 7,605 to just over 1,000. The best accuracy obtained in the former paper using 7,605 features is 97.3% for malware versus cleanware detection and 97.4% for malware family classification; on the reduced feature set, we obtain a (best) accuracy of 94.6% on the malware versus cleanware test and 94.5% on the malware classification test. An interesting feature of the new tests presented here is the reduction in false negative rates by a factor of about 1/3 when compared with the results of the earlier paper. In addition, the speed with which our tests run is reduced by a factor of approximately 3/5 from the times posted for the original paper. The small loss in accuracy and improved false negative rate along with significant improvement in speed indicate that feature reduction should be further pursued as a tool to prevent algorithms from becoming intractable due to too much data.

Linking identities and viewpoints in home movies based on robust feature matching

The identification of useful structures in home video is difficult because this class of video is distinguished from other video sources by its unrestricted, non edited content and the absence of regulated storyline. In addition, home videos contain a lot of motion and erratic camera movements, with shots of the same character being captured from various angles and viewpoints. In this paper, we present a solution to the challenging problem of clustering shots and faces in home videos, based on the use of SIFT features. SIFT features have been known to be robust for object recognition; however, in dealing with the complexities of home video setting, the matching process needs to be augmented and adapted. This paper describes various techniques that can improve the number of matches returned as well as the correctness of matches. For example, existing methods for verification of matches are inadequate for cases when a small number of matches are returned, a common situation in home videos. We address this by constructing a robust classifier that works on matching sets instead of individual matches, allowing the exploitation of the geometric constraints between matches. Finally, we propose techniques for robustly extracting target clusters from individual feature matches.

The dissimilarity corner detector

We present a corner detector that works by using dissimilarity along the contour direction to detect curves in the image contour. The operator is fast, robust to noise and almost self-thresholding. The standard deviation of the image noise must be specified, but this value is easily measured and the explicit modeling of image noise contributes to the robustness of the operator to noise. We also present a new interpretation of the Kitchen-Rosenfeld corner operator (1982) in which we show that this operator can also be viewed as the second derivative of the image function along the edge direction.

2-D edge feature extraction to subpixel accuracy using the generalized energy approach

Precision edge feature extraction is a very important step in vision, Researchers mainly use step edges to model an edge at subpixel level. In this paper we describe a new technique for two dimensional edge feature extraction to subpixel accuracy using a general edge model. Using six basic edge types to model edges, the edge parameters at subpixel level are extracted by fitting a model to the image signal using least-.squared error fitting technique.

Copy number variation and transposable elements feature in recent, ongoing adaptation at the Cyp6g1 locus

The increased transcription of the Cyp6g1 gene of Drosophila melanogaster, and consequent resistance to insecticides such as DDT, is a widely cited example of adaptation mediated by cis-regulatory change. A fragment of an Accord transposable element inserted upstream of the Cyp6g1 gene is causally associated with resistance and has spread to high frequencies in populations around the world since the 1940s. Here we report the existence of a natural allelic series at this locus of D. melanogaster, involving copy number variation of Cyp6g1, and two additional transposable element insertions (a P and an HMS-Beagle). We provide evidence that this genetic variation underpins phenotypic variation, as the more derived the allele, the greater the level of DDT resistance. Tracking the spatial and temporal patterns of allele frequency changes indicates that the multiple steps of the allelic series are adaptive. Further, a DDT association study shows that the most resistant allele, Cyp6g1-[BP], is greatly enriched in the top 5% of the phenotypic distribution and accounts for ~16% of the underlying phenotypic variation in resistance to DDT. In contrast, copy number variation for another candidate resistance gene, Cyp12d1, is not associated with resistance. Thus the Cyp6g1 locus is a major contributor to DDT resistance in field populations, and evolution at this locus features multiple adaptive steps occurring in rapid succession.

A novel alcohol detector based on ZrO2-doped SnO2 electrospun nanofibers

To overcome the interference of acetone when detecting alcohol, a novel alcohol detector based on zirconia-doped SnO2 nanofibers were fabricated through electrospinning technique and calcination process. The samples have been characterized by scanning electron microscopy, transmission electron microscopy, X-ray diffraction, X-ray photoelectron spectroscopy, and their gas sensing properties have also been investigated. When exposed to alcohol vapor, the nanofibers containing 15 mol% zirconia exhibit the best sensing properties. Moreover, the sensor holds the successful discrimination between acetone and alcohol, which makes our product a good candidate in fabricating highly selective sensors in practice.

Using feature selection for intrusion detection system

A good intrusion system gives an accurate and efficient classification results. This ability is an essential functionality to build an intrusion detection system. In this paper, we focused on using various training functions with feature selection to achieve high accurate results. The data we used in our experiments are NSL-KDD. However, the training and testing time to build the model is very high. To address this, we proposed feature selection based on information gain, which can detect several attack types with high accurate result and low false rate. Moreover, we executed experiments to category each of the five classes (probe, denial of service (DoS), user to super-user (U2R), and remote to local (R2L), normal). Our proposed outperform other state-of-art methods.

Generation of 3D sparse feature models using multiple stereo views

Augmented Reality (AR) renders virtual information onto objects in the real world. This new user interface paradigm presents a seamless blend of the virtual and real, where the convergence of the two is difficult to discern. However, errors in the registration of the real and virtual worlds are common and often destroy the AR illusion. To achieve accurate and efficient registration, the pose of real objects must be resolved in a quick and precise manner.