Search

In communications, innovative paradigm shifts have emerged in integrating various devices into the network to provide advanced and intelligent services. However, various security threats may occur that may not always be detected using traditional cryptographic techniques. Secure authentication is of paramount importance in modern wireless systems. This paper focusses on robust authentication in a time-varying communication environment where conventional authentication mechanisms are severely limited. We propose an Adaptive Neural Network (ANN) as an intelligent authentication process to improve detection accuracy. Specifically, a Data-Adaptive Matrix (DAM) is designed to track time-varying channel features. By utilizing a convolutional neural network as an intelligent authenticator, the proposed approach integrates deep feature extraction and attack detection, hence, leading to effective physical layer security. To evaluate the system, the ANN is prototyped on a universal software radio peripheral (USRP) and its authentication performance is evaluated in a conference room environment. Experimental results show that the ANN is effective in tackling the challenges of physical layer authentication under interference conditions, and is effective in time-varying environments.

The performance of classical security authentication models can be severely affected by imperfect channel estimation as well as time-varying communication links. The commonly used approach of statistical decisions for the physical layer authenticator faces significant challenges in a dynamically changing, non-stationary environment. To address this problem, this paper introduces a deep learning-based authentication approach to learn and track the variations of channel characteristics, and thus improving the adaptability and convergence of the physical layer authentication. Specifically, an intelligent detection framework based on a Convolutional-Long Short-Term Memory (Convolutional-LSTM) network is designed to deal with channel differences without knowing the statistical properties of the channel. Both the robustness and the detection performance of the learning authentication scheme are analyzed, and extensive simulations and experiments show that the detection accuracy in time-varying environments is significantly improved.

Wireless networks strive to integrate information technology into every corner of the world. This openness of radio propagation is one reason why holistic wireless security mechanisms only rarely enter the picture. In this paper, we propose a physical (PHY)-layer security authentication scheme that takes advantage of channel randomness to detect spoofing attacks in wireless networks. Unlike most existing authentication techniques that rely on comparing message information between the legitimate user and potential spoofer, our proposed authentication scheme uses a Gaussian mixture model (GMM) to detect spoofing attackers. Probabilistic models of different transmitters are used to cluster messages. Furthermore, a 2-D feature measure space is exploited to preprocess the channel information. Training data for a spoofer operating through an unknown channel, a pseudo adversary model is developed to enhance the spoofing detection performance. Monte Carlo simulations are used to evaluate the detection performance of the GMM-based PHY-layer authentication scheme. The results show that the probability of detecting a spoofer is higher than that obtained using similar approaches.

Physical (PHY)-layer security has received considerable interest as a way to safeguard data confidentiality and achieve security and privacy in wireless networks. Authentication between two devices is a challenging problem. In this paper, a machine learning algorithm is proposed to detect and identify rogue transmitters relying on a low-dimensional channel feature vector that is obtained by the Karhunen-Loeve transform (KLT). Specifically, a Linde-Buzo-Gray algorithm is designed for improving the reliability and robustness of the proposed scheme, where a Gaussian Mixture Model (GMM) is employed to learn and track the changes of physical layer properties. Simulation results demonstrate that the proposed authentication scheme achieves a higher spoofing detection rate compared to other existing methods.

Video-based activity and behavior analysis of mice has garnered wide attention in biomedical research. Animal facilities hold large numbers of mice housed in “home-cages” densely stored within ventilated racks. Automated analysis of mice activity in their home-cages can provide a new set of sensitive measures for detecting abnormalities and time-resolved deviation from the baseline behavior. Large-scale monitoring in animal facilities requires minimal footprint hardware that integrates seamlessly with the ventilated racks. The compactness of hardware imposes the use of fisheye lenses positioned in close proximity to the cage. In this paper, we propose a systematic approach to accurately estimate the 3D pose of the mouse from single-monocular fisheye-distorted images. Our approach employs a novel adaptation of a structured forest algorithm. We benchmark our algorithm against existing methods. We demonstrate the utility of the pose estimates in predicting mouse behavior in a continuous video.

This paper describes a system of practical technologies to implement an illumination-robust, consumer grade biometric system based on face recognition to be used in the automotive market. Most current face recognition systems are compromised in accuracy by ambient illumination changes. Outdoor applications including vehicle personalization pose an especially challenging environment for face recognition. The point of this research is to investigate practical face recognition used for identity management in order to minimize algorithmic complexity while making the system robust to ambient illumination changes. First, a frame differencing method is presented with an active near-infrared illumination control that produces images independent of the ambient illumination. Second, an end-to-end face recognition system is presented including foreground/background segmentation, motion detection, face detection, motion interpolation, pose clustering and face recognition modules. It is shown that the frame differencing method makes the modules more robust to the ambient illumination variation. Vehicular application videos were taken in extremely challenging outdoor illumination and shadowing conditions and used to test each module. Finally, extensive test results of vehicular scenario are provided to evaluate the end-to-end systems1.

A new training algorithm for neural networks in binary classification problems is presented. It is based on the minimization of an estimate of the Bayes risk by using Parzen windows applied to the final one-dimensional nonlinear transformation of the samples to estimate the probability of classification error. This leads to a very general approach to error minimization and training, where the risk that is to be minimized is defined in terms of integrated one-dimensional Parzen windows, and the gradient descent algorithm used to minimize this risk is a function of the window that is used. By relaxing the constraints that are typically applied to Parzen windows when used for probability density function estimation, for example by allowing them to be non-symmetric or possibly infinite in duration, an entirely new set of training algorithms emerge. In particular, different Parzen windows lead to different cost functions, and some interesting relationships with classical training methods are discovered. Experiments with synthetic and real benchmark datasets show that with the appropriate choice of window, fitted to the specific problem, it is possible to improve the performance of neural network classifiers over those that are trained using classical methods. (C) 2017 Elsevier Ltd. All rights reserved. This work was partly supported by Grant TEC-2015-67719-P “Macro-ADOBE” (Spain MINECO/EU FSE, FEDER), and network TIN 2015-70808-REDT, “DAMA” (MINECO) (M. Lázaro and A.R. Figueiras-Vidal), and by Prof. Monson Hayes’ Banco de Santander-UC3M Chair of Excellence, 2015.

Recognizing handwritten equations is a challenging problem, and even more so when they are written in a classroom environment. However, since videos of the handwritten text and the accompanying audio refer to the same content, a combination of video and audio based recognition has the potential to significantly improve the recognition accuracy. In this paper, using a combination of video and audio based recognizers, we focus on improving the character recognition accuracy for handwritten mathematical content in videos using audio and propose an end-to-end recognition system. The system includes components for video preprocessing, selecting the characters that may benefit from audio-video based combination, establishing a correspondence between handwritten and the spoken content, and finally combining the recognition results from the audio and video based recognizers. The current implementation of the system makes use of a modified open source text recognizer and a commercially available phonetic word spotter. For evaluation purposes, we use videos recorded in a classroom-like environment and our experiments demonstrate the significant improvements in character recognition accuracy that can be achieved using our techniques.

Video-based activity and behavior analysis for mice has garnered wide attention in biomedical research. Animal facilities hold large numbers of mice housed in ‘home-cages’ densely stored within ventilated racks. Automated analysis of mice activity in their home-cages can provide a new set of sensitive measures for detecting abnormalities and time-resolved deviation from baseline behavior. Large scale monitoring in animal facilities requires minimal footprint hardware that integrates seamlessly with the ventilated racks. Compactness of hardware imposes use of fisheye lenses positioned in close proximity to the cage. In this paper, we estimate the 3D pose of a mouse from fisheye distorted monocular monochromatic images using a novel adaptation of a structured forests algorithm. The method utilizes classification decision trees leveraging their versatility to store arbitrary information in the leaf-nodes. During training, the samples arriving at each node are mapped from continuous pose space to discrete class labels such that similar poses are grouped in the same class. The node splitting function is trained by optimizing a classification objective function rather than a high-dimensional regression one. The leaf-nodes store the pose parameters for the set of samples reaching the node. A prediction model preserving the structural relationship of the pose is formed based on the samples in the leaf-nodes. We apply the method to what we believe is the first known training set for 3D recovery of mouse key points from monocular images. We compare the results of our approach to those obtained via standard regression techniques.

We address the challenging problem of estimating 3D position of laboratory mouse key-points from monocular images acquired through a fisheye lens positioned very close to the mouse ‘home-cage’. This video acquisition system optical configuration is used when designing practical compact systems for large scale automated monitoring of mice in animal facility ventilated racks. The space constraints in the ventilated racks necessitate the use of fisheye lenses to ensures a more complete view of the home-cage. We extend a cascaded pose regression (CPR) algorithm that has proven to be successful for 2D pose estimation and introduce novel enhancements to the algorithm. Our 3D CPR algorithm reliably generates pose estimates starting from a rough initial pose estimate.

In recent years, researchers and laboratory support companies have recognized the utility of automated profiling of laboratory mouse activity and behavior in the home-cage. Video-based systems have emerged as a viable solution for non-invasive mouse monitoring. Wider use of vision systems for ethology studies requires the development of scalable hardware seamlessly integrated with vivarium ventilated racks. Compact hardware combined with automated video analysis would greatly impact animal science and animal-based research. Automated vision systems, free of bias and intensive labor, can accurately assess rodent activity (e.g., well-being) and behavior 24-7 during research studies within primary home-cages. Scalable compact hardware designs impose constraints, such as use of fisheye lenses, placing greater burden (e.g., distorted image) on downstream video analysis algorithms. We present novel methods for analysis of video acquired through such specialized hardware. Our algorithms estimate the 3D pose of mouse from monocular images. We present a thorough examination of the algorithm training parameters’ influence on system accuracy. Overall, the methods presented offer novel approaches for accurate activity and behavior estimation practical for large-scale use of vision systems in animal facilities.

A new night-time lane detection system and its accompanying framework are presented in this paper. The accompanying framework consists of an automated ground truth process and systematic storage of captured videos that will be used for training and testing. The proposed Advanced Lane Detector 2.0 (ALD 2.0) is an improvement over the ALD 1.0 or Layered Approach with integration of pixel remapping, outlier removal, and prediction with tracking. Additionally, a novel procedure to generate the ground truth data for lane marker locations is also proposed. The procedure consists of an original process called time slicing, which provides the user with unique visualization of the captured video and enables quick generation of ground truth information. Finally, the setup and implementation of a database hosting lane detection videos and standardized data sets for testing are also described. The ALD 2.0 is evaluated by means of the user-created annotations accompanying the videos. Finally, the planned improvements and remaining work are addressed.

Consumer video surveillance systems often suffer from bad weather conditions, observed objects lose visibility and contrast due to the presence of atmospheric haze, fog, and smoke. In this paper, we present an image defogging algorithm with color correction in the HSV color space for video processing. We first generate a modified transmission map of the image segmentation using multiphase level set formulation from the intensity (V) values. We also estimate atmospheric light in the intensity (V) values. The proposed method can significantly enhance the visibility of foggy video frames using the estimated atmospheric light and the modified transmission map. Another contribution of the proposed work is the compensation of color distortion between consecutive frames using the temporal difference ratio of HSV color channels. Experimental results show that the proposed method can be applied to consumer video surveillance systems for removing atmospheric artifacts without color distortion.

In this paper, we present a novel high dynamic range (HDR) imaging method using a single image. The existing multiple image-based HDR methods work only on condition that there is no camera and object movement when acquiring multiple, differently exposed low dynamic range (LDR) images. To overcome such an unrealistic restriction, we make three LDR images from a single input image using local histogram stretching. An edge-preserving spatially adaptive denoising method is also proposed to suppress the noise that is amplified in the histogram stretching process. Because the proposed method self-generates three histogram-stretched LDR images from a single input image, ghost artifacts that are the result of the relative motion between the camera and objects during exposure time, are inherently removed. Therefore, the proposed method can be applied to mobile imaging devices such as a mobile phone camera and a consumer compact camera to provide the ghost artifacts free HDR function in the form of either embedded or post-processing software.

In this paper, we describe how tablet PCs are being used at Georgia Tech Savannah (GTS) to improve student learning in a distributed classroom environment. The Tablet PC is an attractive technology for use in synchronous distributed learning environments because of its mobility, and its ability to not only serve as an effective note taking device but also as a high-resolution course content viewing device and a tool for interactive assessments. The research questions addressed here are: (1) "What impact does the Tablet PC have on student perceptions of their engagement in a distributed learning environment?" and (2) "Can the Tablet PC be used to improve student learning in a distributed learning environment?". In this project, the instructor and students were given a Tablet PC to use during the semester, and surveys were administered to evaluate student attitudes about the use of Tablet PC technology as a means of receiving, processing, and learning course material. The significance of this work is that it serves as a case study on the use of Tablet PCs as an effective technology for implementing established educational practices in distributed education environments.

This paper describes an illumination and pose invariant face recognition system that is intended to be used in the automotive market for vehicle personalization. Near-infrared frame differencing improves the robustness to the outdoor illumination conditions. And we introduce the video-based recognition with pose clustering for pose invariant face recognition. We have collected large video dataset of drivers under various extreme outdoor illumination conditions. Experimental results show that the proposed method achieves high recognition rate with near-infrared frame differencing and pose clustering.

The Georgia Institute of Technology in Atlanta is in a unique position of having a campus located in the heart of the Lorraine region of France in the city of Metz. Since the number of faculty at GTL is limited, some courses must be offered in non-traditional formats. Due to the problems involved in shipping video tapes across the Atlantic, the Internet was considered as an alternative delivery vehicle. Initially driven by a very real and pressing need to deliver courses in a timely and effective manner across the Atlantic, Internet course delivery is now finding its way into many other programs at Georgia Tech. In this paper, we describe some of our experiences with Internet course development and delivery, and present a set of tools that we have developed for placing courses on the web easily. We will also present some of the problems and limitations with on-line course production and delivery, and speculate on the future of distance education.

In this paper, a real time dynamic gesture recognition method based on a modified Levenshtein distance is proposed. The method addresses the issues faced in dynamic gesture recognition methods that are due to gesture variability within the class due to differences in the gesture speed. Since the gesture speed and duration will vary from one gesture to another, or from one person to the next, without taking this variability into account there will be an increase in the number of recognition errors. Here, we use a modified Levenshtein distance to represent each gesture class by one gesture exemplar and no restrictions on the gesture speed is necessary. Experiments demonstrate real-time gesture recognition rates of 93.9% for gestures of variable speed.

This paper is presents an artwork that is concerned with the interactions among people rather than the interaction between an audience and the artwork. We visualize the physical motion variations from the interactions among different participants using Kinect-based depth estimation and video tracking algorithms. The proposed work can visualize the affective experiences based on the physical distance between participants. We also provide experiences in which a participant becomes a part of the artwork in the form of both shape and interface. The body of a participant plays an important role in communicating and interacting with other participant and the artwork itself.

In this paper, we consider the problem of recognizing handwritten mathematical content from classroom videos. Since the handwritten text and the accompanying audio refer to the same mathematical characters and symbols, a combination of video and audio based recognizers has the potential to significantly increase the recognition accuracy compared to that of the individual recognizers. In this paper, we propose a novel multi-step technique for combining the output of the video and the audio based recognizers. Initial recognition results from a video based recognizer and a speech recognizer, operating independently on the handwritten and the spoken content from a classroom video, are combined with a base mathematical speech grammar to arrive at a constrained speech grammar that is specific to the content being recognized. The constrained speech grammar is then used by the speech recognizer to generate the final character recognition results. A subsequent layout analysis step, which makes used of audio cues and X-Y cuts based method, is used to arrive at the final recognized content. Experiments conducted using videos recorded in a classroom like environment are used to demonstrate the significant improvement in recognition accuracy that can be achieved using our technique.

In this paper, we present an image enhancement algorithm based on the wavelength-adaptive image formation model and geometric classification for defogging UAV images. We first generate a labeled image using geometric class-based segmentation. We then generate a modified transmission map based on the wavelength-adaptive image formation model with scattering coefficients in the labeled image. We also estimate the atmospheric light from the modified transmission map instead of simply choosing the brightest pixel. The proposed method can significantly enhance the visibility of foggy UAV images compared with existing monochrome model-based defogging method. The proposed algorithm can enhance the visibility by removing atmospheric degradation factor in airborne images acquired by aerial platforms such as satellite, airplane, and UAV under critical weather conditions such as haze, fog, and smoke.

In this paper, we present a dual off-axis color filtered aperture (DCA)-based computational imaging system for depth estimation in a single-camera framework. The DCA has one primary (red) color and its complementary (cyan) color filtered apertures located off the optical axis to generate misalignment between color channels depending on the distance of a region-of-interest from the camera. Disparity of color shifting values (CSVs) between color channels in the image acquired by the DCA-based imaging system is estimated using L1-norm minimization of energy functional considering both brightness and gradient constancies and only translation of x-axis. The two data terms both red to green (R-G) and red to blue (R-B) are combined in a single data term. The proposed imaging system can be implemented by simply inserting an appropriately resized DCA into any general optical system. Experimental results show that the proposed DCA can estimate the distance of the scene from the camera in the single-camera framework.

In this paper, we present a novel contrast enhancement method for backlit images that consists of three steps: i) computation of the transmission coefficients using the dark channel prior, ii) generation of multiple images having different exposures based on the transmission coefficients, and iii) image fusion. Compared to global intensity transformation methods and spatially invariant contrast enhancement algorithms, our approach first extracts under-exposed regions using the dark channel prior map, and then performs spatially adaptive contrast enhancement. As a result, the contrast of the image is increased, especially for backlit scenes and those with very wide dynamic range, while still preserving image details and color.

When a camera is modified by placing two or more displaced apertures with color filters within the imaging system, it is possible to estimate the distances of objects from the camera and to create 3-d images. In this paper, we develop the key equations necessary to estimate the distance of an object and discuss the feasibility of such a system for distance estimation in applications such as robot vision, human computer interfaces, intelligent visual surveillance, 3-d image acquisition, and intelligent driver assistance systems. In particular, we discuss how accurately these distances may be estimated and describe how distance estimation may be performed in real-time using an appropriately modified video camera.

In this paper, we present an efficient backlit region detection and enhancement method with low-computational cost for consumer mobile imaging devices. The proposed approach partitions an input image into multiple rectangular blocks, and uses a fuzzy logic-based optimal threshold to classify each block as backlit, background, or ambiguous. Backlit blocks are then processed using a contrast enhancement algorithm, whereas those block that are classified as ambiguous are further partitioned into four subblocks, and each subblock goes through further classification. The major advantage of the proposed method is computational efficiency and improved quality with minimized blocking artifacts compared with other block-based approaches.

In this paper, we present a robust motion-based object detection system that corrects for the motion of an unstable camera. Assuming that the global camera motion may be modeled as an affine transform of the image between two successive frames, the proposed method is able to correct for camera motion using an elastic registration algorithm (ER). The local motion is then estimated from a current image and affine-transformed previous image. Finally, object regions are detected using the estimated local motions. Experimental results show that the proposed system is able to robustly detect moving objects in unstable imaging environment for consumer surveillance systems.

In this paper, we present an efficient moving object segmentation method using motion orientation histogram (MOH) in consideration of variable block-based hardware implementation. In pursuit of both efficiency and reliability, each block motion vector is quantized into one of eight representative orientations. Given a set of motion vectors estimated from regularly divided basic blocks, we adaptively partition the blocks based on entropy for increasing the reliability of estimated motions. We then compute motion orientation histogram (MOH) from appropriately partitioned blocks and quantize them into eight representative orientations. Finally, we decide the object's motion using the quantized orientation of motion and error compensation. Experimental results show that the proposed method can be embedded in an image signal processing (ISP) chip for high-level image processing functions such as object tracking and behavior analysis in consumer surveillance systems.

In this paper, we present an adaptive background generation method for automatic selection of initial object regions, which realizes simultaneous object detection and depth estimation using multiple color-filter aperture (MCA) camera. Since the conventional background generation method does not fit the depth estimation using the MCA camera, we propose a novel color-based background generation method which can reduce interference in the object region for stable depth estimation. For efficient estimation of color shifting vectors in the extracted object region, a simplified elastic registration (ER) algorithm is used. The proposed simplified method is essential factor to realize realtime depth estimation and tracking, which is the primary condition for consumer applications. Finally, the object distance is determined by using the relationship between the pre-specified distance transformation function and the estimated shifting vectors of the corresponding object region. Although traditional depth estimation methods generally use dual cameras for stereo vision, the proposed method uses only a single camera for both object detection and depth estimation. Experimental results show that the proposed MCA camera-based object detection system can be used in a variety of consumer surveillance systems such as intelligent transport systems, 3D-based cameras and advanced safety vehicles.

Recognizing handwritten mathematical content is a challenging problem, and more so when such content appears in classroom videos. However, given the fact that in such videos the handwritten text and the accompanying audio refer to the same content, a combination of a video and an audio based recognizer has the potential to significantly improve the content recognition accuracy. In this paper, using a combination of video and audio based recognizers, we focus on improving the character recognition accuracy in such videos and propose: (1) synchronization techniques for establishing a correspondence between the handwritten and the spoken content, and (2) combination techniques for combining the outputs of the video and audio based recognizers. The current implementation of the system makes use of a modified open source text recognizer and a commercially available phonetic word-spotter. For evaluation purposes, we use videos recorded in a classroom-like environment and our experiments demonstrate the significant improvements (≈ 24% relative increase as compared to the baseline video based recognizer) in character recognition accuracy that can be achieved using our techniques.

Recognizing handwritten mathematical content in classroom videos poses a range of interesting challenges. In this paper, we focus on improving the character recognition accuracy in such videos using a combination of video and audio based text recognizers. We propose a two step assembly consisting of a video text recognizer (VTR) as the primary character recognizer and an audio text recognizer (ATR) for disambiguating, if needed, the output of the VTR. We propose techniques for (1) detecting ambiguity in the output of the VTR so that a combination with the ATR may be triggered only for ambiguous characters, (2) synchronizing the output of the two recognizers for enabling combination, and (3) combining the options generated by the two recognizers using measurement and rank based methods. We have implemented the system using an open source implementation of a character recognizer and a commercially available phonetic word-spotter. Through experiments conducted using video recorded in a classroom-like environment, we demonstrate the improvement in the character recognition accuracy that can be achieved using our approach.

In classroom videos, recognizing mathematical content, handwritten on the whiteboard presents a unique opportunity in the form of audio content spoken by the instructor. This recognized audio content can be used to improve the character recognition accuracy by providing evidence in corroboration or contradiction of the output options generated by the primary, video based recognizer. However, such audio-video based disambiguation also has the potential to introduce errors in what may have been the correct output from the video based recognizer. In this paper, we focus on improving the character recognition accuracy by developing ambiguity detection methods that can be used to determine the set of potentially incorrect outputs from the video based recognizer and, for each such output, determining the subset of possibly correct output options that must be forwarded for audio-video based character disambiguation. In this paper, we propose, implement and evaluate a number of such ambiguity detection methods.

In this paper, we propose a new methodology for detecting lane markers that exploits the parallel nature of lane boundaries on the road. First, the input image is pre-processed and filtered to detect lane marker features. Then, using the Polar Randomized Hough Transform that is introduced in this paper, lines are fitted through the detected features and the orientation of each line is evaluated. By finding near parallel lines separated by a constraint specified distance, false signalling caused by artifacts in the image is greatly reduced. The proposed system was tested using a real world driving videos and showed good results despite the presence of neighboring vehicles, shadows, and irregularities on the road surface.

In this paper, we present a new multi camera approach to Lane Departure Warning (LDW). Upon acquisition, the captured images are transformed to a bird's-eye view using a modified perspective removal transformation. Then, camera calibration is used to accurately determine the position of the two cameras relative to a reference point. Lane detection is performed on the front and rear camera images which are combined using data fusion. Finally, the distance between the vehicle and adjacent lane boundaries is determined allowing to perform LDW. The proposed system was tested on real world driving videos and shows good results when compared to ground truth.

Lane detection is an important application of driver assistance. In this paper, a new technique for detecting lane markers that is able to cope with many complex conditions is presented. Some of these conditions include dynamic illumination, scattered shadows, and the presence of neighboring vehicles to name a few. The input image is first pre-processed with a perspective removal transformation followed by a color space conversion. Then, the core elements of the proposed technique consisting of template matching, lane region merging, elliptical projections, and parametric tracking are explained. A formal error metric used in performance evaluation is also introduced. Finally, quantitative analyses show that the developed system performs well in real-world driving conditions with variations in illumination, traffic, and road surface quality.

In this paper, we present a new methodology for detecting lane markers that is able to withstand many challenging situations like scattered shadows, illumination changes, and presence of neighboring vehicles to name a few. At first, the input image undergoes a perspective removal followed by a color space conversion. Then, the core elements consisting of template matching, lane region merging, and elliptical projections are explained. Finally, the developed system showed good results when tested on 15 minutes of real-world driving videos containing variations in illumination, traffic, and road surface conditions.

In this document, a new and efficient method to specify the ground truth locations of lane markers is presented. The method comprises of a novel process called Time-Slicing that provided the user with a unique visualization of the video. Coupled with automation via spline interpolation, the quick generation of necessary ground truth information is achieved. Videos recorded from a vehicle while driving on local city roads and highways are marked with ground truth information for use in testing. The performance of a variety of lane detection systems is compared to the ground truth and the error is computed for each system. Finally, quantitative analysis shows that the reference lane detection system presented in [1] produces the most accurate lane detections which is depicted by the smallest error.

Lane detection is an important element of most driver assistance applications. A new lane detection technique that is able to withstand some of the common issues like illumination changes, surface irregularities, scattered shadows, and presence of neighboring vehicles is presented in this paper. At first, inverse perspective mapping and color space conversion is performed on the input image. Then, the images are cross-correlated with a collection of predefined templates to find candidate lane regions. These regions then undergo connected components analysis, morphological operations, and elliptical projections to approximate positions of the lane markers. The implementation of the Kalman filter enables tracking lane markers on curved roads while RANSAC helps improve estimates by eliminating outliers. Finally, a new method for calculating errors between the detected lane markers and ground truth is presented. The developed system showed good performance when tested with real-world driving videos containing variations in illumination, road surface, and traffic conditions.

This paper introduces a generalized symbolic signal representation capable of capturing the physical appearance of process monitoring signals. The signals are decomposed into simpler subcomponents described by different parametric models and combined under an object-oriented hybrid signal model. These subcomponents are modeled by a piecewise spline function approximation. This model represents the signal patterns in purely mathematical terms and can be viewed as an extension of syntactic signal modeling approaches used in electrocardiogram (ECG) analysis. Furthermore, a robust event-synchronous system capable of transcribing given signals into the proposed symbolic representation is described. Using event-synchronous analysis makes the system specifically useful for process monitoring applications.

In a previous paper, a simple approach to lane detection using the Hough transform and iterated matched filters was described [1]. This paper extends this work by incorporating an inverse perspective mapping to create a bird's-eye view of the road, applying random sample consensus to help eliminate outliers due to noise and artifacts in the road, and a Kalman filter to help smooth the output of the lane tracker.

Distributed education is becoming an integral part of the programs at Georgia Tech. However, the traditional mode of synchronous distributed learning (DL) classes has typically suffered from several challenges that hinder effective teaching and learning practices. This paper is a case study that outlines work on a project to implement Tablet PCs into the design of DL courses to enhance their effectiveness in synchronous deployment. Feedback from students and faculty were collected as a means of assessing the potential impact of the Tablet PC modality in synchronous DL environments. Results indicate that this modality has several distinct advantages over more traditional DL modalities from the perspective of both students and faculty.

A new night-time lane detection system that extends the idea of a Layered Approach [1] is presented in this document. The extension includes the incorporation of (1) Inverse Perspective Mapping (IPM) to generate a bird’s-eye view of the road surface, (2) application of Random Sample Consensus (RANSAC) to rid outliers from the data, and (3) Kalman filtering to smooth the output of the lane tracker. Videos of driving scenarios on local city roads and highways were used to test the new system. Quantitative analysis shows higher accuracy in detecting lane markers in comparison to other approaches.

Face images that are captured by surveillance cameras usually have a very low resolution, which significantly limits the performance of face recognition systems. In the past, super-resolution techniques have been proposed to increase the resolution by combining information from multiple images. These techniques use super-resolution as a preprocessing step to obtain a high-resolution image that is later passed to a face recognition system. Considering that most state-of-the-art face recognition systems use an initial dimensionality reduction method, we propose to transfer the super-resolution reconstruction from pixel domain to a lower dimensional face space. Such an approach has the advantage of a significant decrease in the computational complexity of the super-resolution reconstruction. The reconstruction algorithm no longer tries to obtain a visually improved high-quality image, but instead constructs the information required by the recognition system directly in the low dimensional domain without any unnecessary overhead. In addition, we show that face-space super-resolution is more robust to registration errors and noise than pixel-domain super-resolution because of the addition of model-based constraints.

The paper describes an application of practical technologies to implement a low cost, consumer grade, single chip biometric system based on face recognition using infra-red imaging. The paper presents a system that consists of three stages that contribute in the face detection and recognition process. Each stage is explained with its individual contribution alongside results of tests performed for that stage. The system shows a high recognition rate when full frontal face images are fed to the system. The paper further discusses the application based approach in the automotive world with plans for further study. Recognition rates of the overall system are also presented.

The objective of this work is to provide a simple and efficient method to detect human faces in the compressed domain. We separate skin regions from non-skin regions using the Bayesian decision rule with a Gaussian mixture model (GMM). Next, we detect the location of a face using a deformable template that considers shape and orientation of the face. Then, face candidates are verified by using the second moments computed directly from the DCT coefficients in the face region. In particular, the coefficients are manipulated differently according to the orientation of the face. Good results have been obtained for a large variety of video sequences. This algorithm can be applied to JPEG images as well, without any modification.

Many techniques are available for the estimation of the power spectrum of a stationary random process. While power spectrum estimation is a problem which falls within the domain of signal processing, the problem of inferring information falls within the domain of artificial intelligence (AI). With a wide variety of different types of power spectrum estimation techniques to choose from, an equally wide range of differing spectral estimates may be produced. Each estimate, however, may be used to infer information about the time series. By defining an appropriate knowledge base, a system is being developed to infer information from power spectrum estimates. This system combines the estimates produced by a variety of current spectrum estimation techniques in order to formulate a composite spectral estimate.

In this paper, the application of a constrained iterative deconvolution algorithm to the problem of enhancing noncoherent radar data is described. It is shown that noncoherent reconstruction is theoretically not unique. Despite this, an approximate convolutional model of the observed data is established and it is seen that reasonable results are observed. A key feature of the algorithm is the use of a median filter and threshold to segment the data into two classes, potential point targets and background, which are deconvolved separately subject to different nonlinear constraints. The performance of the algorithm is demonstrated using real air-to-ground millimeter-wave noncoherent radar data.

This paper proposes a novel system that can automatically create an optimal and nonrepetitive summarization and support different user requirements for video browsing and content overview by outputting both the optimal set of key frames and a summarized version of the original video with the user-specified time length. Comparing our approach to video abstraction with another algorithm, we demonstrate that our approach is fast and produces an effective video summary.

Face recognition is an important part of today's emerging biometrics and video surveillance markets. Recent years have witnessed an exploding interest in the development of face recognition algorithms and products. Currently, face recognition systems are usually implemented on general purpose processors. As face recognition algorithms move from research labs to the real world, power consumption and cost become critical issues. This motivates searching for implementations using a digital signal processor (DSP). Our goal is to explore the feasibility of implementing DSP-based face recognition systems. To achieve this goal, we implement a fully automatic face recognition system on Texas Instruments' TMS320C6416 DSP, profile performance, and analyze opportunities for optimization. The results of our experiments demonstrate that a generic C implementation with a modest C level optimization effort results in a face recognition software prototype that has low CPU and memory requirements. Hence, it appears that well-optimized face recognition implementations on DSPs can be an effective choice for embedded face recognition products.

Face images that are captured by surveillance cameras usually have a very low resolution, which significantly limits the performance of face recognition systems. In the past, super-resolution techniques have been proposed that attempt to increase the resolution by combining information from multiple images. These techniques use super-resolution as a preprocessing system to obtain a high resolution image that can later be passed to a face recognition system. Considering that most state-of-the-art face recognition systems use an initial dimensionality reduction method, we propose embedding the super-resolution algorithm into the face recognition system so that super-resolution is not performed in the pixel domain, but is instead performed in a reduced dimensional domain. The advantage of such an approach is a significant decrease in the computational complexity of the super-resolution algorithm because the algorithm no longer tries to construct a visually improved high quality image, but instead constructs the information required by the recognition algorithm directly in the lower dimensional domain without any unnecessary overhead.