A locally-connected neural network for fingerprint recognition
Fingerprint recognition is typically done by matching graphs or templates of extracted minutia from whole fingerprint images.
SEARCH · Engineering Papers
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Fingerprint recognition is typically done by matching graphs or templates of extracted minutia from whole fingerprint images.
We utilize the context information of other regions in hierarchical image segmentation to learn new regions affinities. It is well known that a single choice of quantization of an image space is highly unlikely to be a common optimal quantization level for all categories. Each level of quantization has its own benefits. Therefore, we utilize the hierarchical information among different quantizations as well as spatial proximity of their regions. The proposed affinity learning takes into account higher order relations among image regions, both local and long range relations, making it robust to instabilities and errors of the original, pairwise region affinities. Once the learnt affinities are obtained, we use a standard image segmentation algorithm to get the final segmentation. Moreover, the learnt affinities can be naturally unutilized in interactive segmentation. Experimental results on Berkeley Segmentation Dataset and MSRC Object Recognition Dataset are comparable and in some aspects better than the state-of-art methods.
Peptidoglycan recognition proteins (PGRPs) recognize bacteria through their unique cell wall constituent, peptidoglycans (PGs). PGRPs are conserved from insects to mammals and all function in antibacterial defense. In the tobacco hornworm Manduca sexta, PGRP1 and microbe binding protein (MBP) interact with PGs and hemolymph protease-14 precursor (proHP14) to yield active HP14. HP14 triggers a serine protease network that produces active phenoloxidase (PO), Spätzle, and other cytokines to stimulate immune responses. PGRP1 binds preferentially to diaminopimelic acid (DAP)-PGs of Gram-negative bacteria and Gram-positive Bacillus and Clostridium species than Lys-PGs of other Gram-positive bacteria. In this study, we synthesized DAP- and Lys-muramyl pentapeptide (MPP) and monitored their associations with M. sexta PGRP1 by surface plasmon resonance. The K d values (0.57 μM for DAP-MPP and 45.6 μM for Lys-MPP) agree with the differential recognition of DAP- and Lys-PGs. To reveal its structural basis, we produced the PGRP1 in insect cells and determined its structure at a resolution of 2.1 Å. The protein adopts a fold similar to those from other PGRPs with a classical L-shaped PG-binding groove. A unique loop lining the shallow groove suggests a different ligand-binding mechanism. In summary, this study provided new insights into the PG recognition by PGRPs, a critical first step that initiates the serine protease cascade.
The paper describes the problem of finding the likelihood ratio for specific problem of stochastic system recognition in continuous time to member functions within continuous-discrete time, which depend not only on current, but also on arbitrary numbers of previous non-observable process values.
Description, recognition and analysis biological images plays an important role for human to describe and understand the related biological information. The color images are separated by color reduction. A new and efficient linearization algorithm is introduced based on some criteria of difference chain code. A series of critical points is got based on the linearized lines. The series of curvature angle, linearity, maximum linearity, convexity, concavity and bend angle of linearized lines are calculated from the starting line to the end line along all smoothed contours. The useful method can be used for shape description and recognition. The analysis, decision, classification of the biological images are based on the description of morphological structures, color information and prior knowledge, which are associated each other. The efficiency of the algorithms is described based on two applications. One application is the description, recognition and analysis of color flower images. Another one is related to the dynamic description, recognition and analysis of cell-cycle images.
A method is proposed for the capture of an object on the basis of fuzzy information regarding the form and physical characteristics of its surface. Exponential or robust filters may be employed. Decisions rules are formulated regarding the tactile recognition of target points on the surface. The condition for successful capture of the object is presented. An algorithm is proposed for the formulation of a target point on the motion plane of active capture elements.
Nucleotide-binding domain leucine-rich repeat-containing receptors (NLRs) in plants can detect avirulence (AVR) effectors of pathogenic microbes. The Mildew locus a (Mla) NLR gene has been shown to confer resistance against diverse fungal pathogens in cereal crops. In barley, Mla has undergone allelic diversification in the host population and confers isolate-specific immunity against the powdery mildew-causing fungal pathogen Blumeria graminis forma specialis hordei (Bgh). We previously isolated the Bgh effectors AVR A1 , AVR A7 , AVR A9 , AVR A13 , and allelic AVR A10 /AVR A22 , which are recognized by matching MLA1, MLA7, MLA9, MLA13, MLA10 and MLA22, respectively. Here, we extend our knowledge of the Bgh effector repertoire by isolating the AVR A6 effector, which belongs to the family of catalytically inactive RNase-Like Proteins expressed in Haustoria (RALPHs). Using structural prediction, we also identified RNase-like folds in AVR A1 , AVR A7 , AVR A10 /AVR A22 , and AVR A13 , suggesting that allelic MLA recognition specificities could detect structurally related avirulence effectors. To better understand the mechanism underlying the recognition of effectors by MLAs, we deployed chimeric MLA1 and MLA6, as well as chimeric MLA10 and MLA22 receptors in plant co-expression assays, which showed that the recognition specificity for AVR A1 and AVR A6 as well as allelic AVR A10 and AVR A22 is largely determined by the receptors’ C-terminal leucine-rich repeats (LRRs). The design of avirulence effector hybrids allowed us to identify four specific AVR A10 and five specific AVR A22 aa residues that are necessary to confer MLA10- and MLA22-specific recognition, respectively. This suggests that the MLA LRR mediates isolate-specific recognition of structurally related AVR A effectors. Thus, functional diversification of multi-allelic MLA receptors may be driven by a common structural effector scaffold, which could be facilitated by proliferation of the RALPH effector family in the pathogen genome.
A Medipix2 device was exposed to radioactive sources ({sup 241}Am, {sup 137}Cs and {sup 106}Ru). To test the reliability of track recognition with this device, the activities of the radioactive sources were extracted from the experimental data and compared to the expected activities.
With increasing deployment of satellite imaging systems, only a small fraction of collected data can be subject to expert scrutiny. We present and evaluate a two-tier approach to broad area search for signs of anthropogenic activities in high-resolution commercial satellite imagery. The method filters image information using semantically oriented interest points by combining Harris corner detection and spatial pyramid matching. The idea is that anthropogenic structures, such as rooftop outlines, fence corners, road junctions, are locally arranged in specific angular relations to each other. They are often oriented at approximately right angles to each other (which is known as rectilinearity relation). Detecting the rectilinearity provides an opportunity to highlight regions most likely to contain anthropogenic activity. This is followed by supervised classification of regions surrounding the detected corner points as man-made vs. natural scenes. We consider, in particular, a search for anthropogenic activities in uncluttered areas. In this paper, we proposed and evaluated a two-tier approach to broad area search for signs of anthropogenic activities. Results from experiments on high-resolution ({approx}0.6m) commercial satellite image data showed the potential applicability of this approach and its ability of achieving both high precision and recall rates. The main advantage of combining corner-based cueing with general object recognition is that the incorporation of domain specific knowledge even in its more general form, such as presence of comers, provides a useful cue to narrow the focus of search for signs of anthropogenic activities. Combination of comer based cueing with spatial pyramid matching addressed the issue of comer categorization. An important practical issue for further research is optimizing the balance between false positive and false negative rates. While the results presented in the paper are encouraging, the problem of an automated broad area search for signs of anthropogenic activities remains challenging.
Computer vision has applications in object detection, image recognition and classification, and object tracking. One of the challenges of computer vision is the presence of useful information at multiple distance scales. Filtering techniques may sacrifice details at small scales in order to prioritize the analysis of large-scale features of the image. We present a strategy for coarse-graining multidimensional data while maintaining fine-grained detail for subsequent analysis. The algorithm is based on fixed-size block segmentation in the feature space. We apply this strategy to solve the long-standing challenge of detecting particle trajectories at the Large Hadron Collider in real time.
In this paper we investigate an Intelligent Computer Vision System applied for recognition and classification of commercially available cork tiles. The system is capable of acquiring and processing gray images using several feature generation and analysis techniques. Its functionality includes image acquisition, feature extraction and preprocessing, and feature classification with neural networks (NN). We also discuss system test and validation results from the recognition and classification tasks. The system investigation also includes statistical feature processing (features number and dimensionality reduction techniques) and classifier design (NN architecture, target coding, learning complexity and performance, and training with our own metaheuristic optimization method). The NNs trained with our genetic low-discrepancy search method (GLPtauS) for global optimisation demonstrated very good generalisation abilities. In our view, the reported testing success rate of up to 95% is due to several factors: combination of feature generation techniques; application of Analysis of Variance (ANOVA) and Principal Component Analysis (PCA), which appeared to be very efficient for preprocessing the data; and use of suitable NN design and learning method.
Convolutional neural networks (CNNs) have been widely applied in the computer vision community to solve complex problems in image recognition and analysis. We describe an application of the CNN technology to the problem of identifying particle interactions in sampling calorimeters used commonly in high energy physics and high energy neutrino physics in particular. Following a discussion of the core concepts of CNNs and recent innovations in CNN architectures related to the field of deep learning, we outline a specific application to the NOvA neutrino detector. This algorithm, CVN (Convolutional Visual Network) identifies neutrino interactions based on their topology without the need for detailed reconstruction and outperforms algorithms currently in use by the NOvA collaboration.
Several feature extraction and selection methods for an existing automatic target recognition (ATR) system using JPLs Grayscale Optical Correlator (GOC) and Optimal Trade-Off Maximum Average Correlation Height (OT-MACH) filter were tested using MATLAB. The ATR system is composed of three stages: a cursory region of-interest (ROI) search using the GOC and OT-MACH filter, a feature extraction and selection stage, and a final classification stage. Feature extraction and selection concerns transforming potential target data into more useful forms as well as selecting important subsets of that data which may aide in detection and classification. The strategies tested were built around two popular extraction methods: Principal Component Analysis (PCA) and Independent Component Analysis (ICA). Performance was measured based on the classification accuracy and free-response receiver operating characteristic (FROC) output of a support vector machine(SVM) and a neural net (NN) classifier.
Nondestructive testing and evaluation are under continuous development. Currently researches are concentrated on three main topics: advancement of existing methods, introduction of novel methods and development of artificial intelligent systems for automatic defect recognition (ADR). Automatic defect classification algorithm comprises of two main tasks: creating a defect database and preparing a defect classifier. Here, the database was built using defect features that describe all geometrical and texture properties of the defect. Almost twenty carefully selected features calculated for flaws extracted from real radiograms were used. The radiograms were obtained from shipbuilding industry and they were verified by qualified operator. Two weld defect's classifiers based on artificial neural networks were proposed and compared. First model consisted of one neural network model, where each output neuron corresponded to different defect group. The second model contained five neural networks. Each neural network had one neuron on output and was responsible for detection of defects from one group. In order to evaluate the effectiveness of the neural networks classifiers, the mean square errors were calculated for test radiograms and compared.
An analogy is made and exploited between the recognition of visual objects and language parsing. A subset of regular languages is used to define a one-dimensional 'visual' language, in which the words are translational and scale invariant. This allows an exploration of the viewpoint invariant languages that can be solved by a network of concurrent, hierarchically connected processors. A language family is defined that is hierarchically tiling system recognizable (HREC). As inspired by nature, an algorithm is presented that constructs a cellular automaton that recognizes strings from a language in the HREC family. It is demonstrated how a language recognizer can be implemented from the cellular automaton using a feed-forward network of spiking neurons. This parser recognizes fixed-length strings from the language in parallel and as the computation is pipelined, a new string can be parsed in each new interval of time. The analogy with formal language theory allows inferences to be drawn regarding what class of objects can be recognized by visual cortex operating in purely feed-forward fashion and what class of objects requires a more complicated network architecture.
The development of analysis methods to distinguish potential beyond the Standard Model phenomena in a model-agnostic way can significantly enhance the discovery reach in collider experiments. However, the typical machine learning (ML) algorithms employed for this task require fixed length and ordered inputs that break the natural permutation invariance in collision events. To address this, a semi-supervised anomaly detection tool is presented that takes a variable number of particle-level inputs and leverages a signal model to encode this information into a permutation invariant, event-level representation via supervised training with a Particle Flow Network (PFN). Data events are then encoded into this representation and given as input to an autoencoder for unsupervised ANomaly deTEction on particLe flOw latent sPacE (ANTELOPE), classifying anomalous events based on a low-level and permutation invariant input modeling. Performance of the ANTELOPE architecture is evaluated on simulated samples of hadronic processes in a high energy collider experiment, showing good capability to distinguish disparate models of new physics.
Abstract The reconstruction of the trajectories of charged particles, or track reconstruction, is a key computational challenge for particle and nuclear physics experiments. While the tuning of track reconstruction algorithms can depend strongly on details of the detector geometry, the algorithms currently in use by experiments share many common features. At the same time, the intense environment of the High-Luminosity LHC accelerator and other future experiments is expected to put even greater computational stress on track reconstruction software, motivating the development of more performant algorithms. We present here A Common Tracking Software (ACTS) toolkit, which draws on the experience with track reconstruction algorithms in the ATLAS experiment and presents them in an experiment-independent and framework-independent toolkit. It provides a set of high-level track reconstruction tools which are agnostic to the details of the detection technologies and magnetic field configuration and tested for strict thread-safety to support multi-threaded event processing. We discuss the conceptual design and technical implementation of ACTS, selected applications and performance of ACTS, and the lessons learned.
Abstract At the High Luminosity Large Hadron Collider (HL-LHC), traditional track reconstruction techniques that are critical for physics analysis will need to be upgraded to scale with track density. Quantum annealing has shown promise in its ability to solve combinatorial optimization problems amidst an ongoing effort to establish evidence of a quantum speedup. As a step towards exploiting such potential speedup, we investigate a track reconstruction approach by adapting the existing geometric Denby-Peterson (Hopfield) network method to the quantum annealing framework for HL-LHC conditions. We develop additional techniques to embed the problem onto existing and near-term quantum annealing hardware. Results using simulated annealing and quantum annealing with the D-Wave 2X system on the TrackML open dataset are presented, demonstrating the successful application of a quantum annealing algorithm to the track reconstruction challenge. We find that combinatorial optimization problems can effectively reconstruct tracks, suggesting possible applications for fast hardware-specific implementations at the HL-LHC while leaving open the possibility of a quantum speedup for tracking.