Recovering Intrinsic Images from a Single Image

We present an algorithm that uses multiple cues to recover shading and reflectance intrinsic images from a single image. Using both color information and a classifier trained to recognize gray-scale patterns, each image derivative is classified as being caused by shading or a change in the surface's reflectance. Generalized Belief Propagation is then used to propagate information from areas where the correct classification is clear to areas where it is ambiguous. We also show results on real images.

Inferring 3D Structure with a Statistical Image-Based Shape Model

We present an image-based approach to infer 3D structure parameters using a probabilistic "shape+structure'' model. The 3D shape of a class of objects may be represented by sets of contours from silhouette views simultaneously observed from multiple calibrated cameras. Bayesian reconstructions of new shapes can then be estimated using a prior density constructed with a mixture model and probabilistic principal components analysis. We augment the shape model to incorporate structural features of interest; novel examples with missing structure parameters may then be reconstructed to obtain estimates of these parameters. Model matching and parameter inference are done entirely in the image domain and require no explicit 3D construction. Our shape model enables accurate estimation of structure despite segmentation errors or missing views in the input silhouettes, and works even with only a single input view. Using a dataset of thousands of pedestrian images generated from a synthetic model, we can perform accurate inference of the 3D locations of 19 joints on the body based on observed silhouette contours from real images.

Festival Images: brand image and stakeholders' brand relationship types at the Edinburgh Festival Fringe.

The Edinburgh Festival Fringe (The Fringe) is the largest arts festival in the world and it has inspired the creation of similar festivals world-wide. Since its conception in 1947, the Fringe has demonstrated significant growth in visitor numbers; ticket sales; and its economic contribution. Despite this, the sustainable future of Edinburgh’s festivals is debated as Edinburgh, ‘the Festival City’, faces threats from other festival destinations. Festivals position Edinburgh creatively in contrast to the city’s traditionally perceived image as a cultural-historic centre. Despite this, little research has been undertaken into the creative and cultural significance of Edinburgh’s festivals, including the Fringe. This interdisciplinary research grounded in marketing, tourism, and festival and event management; and underpinned by constructivism, presents an understanding of types of brand relationships that exist between the Fringe and its primary stakeholders. This is achieved through defining both the Fringe brand image and its primary stakeholders; and applying these definitions to the development of a typology of Fringe-stakeholders’ brand relationships. The significance of this study is evident within its topic of inquiry and the research methods applied. In the little-considered arena of arts festivals and their stakeholders, this is the first in-depth study into the Fringe as a festival and festival brand. Within this, the definition of a Fringe brand image contributes to understanding the cultural and creative significance of the Fringe. Furthermore, this research contributes a unique understanding of the types of stakeholders that are engaged with the Fringe. The types of brand relationships that exist between these stakeholders and the Fringe are another significant contribution to knowledge and understanding. While specific to the present context, these findings may prove transferable to further festivals or events, and related areas and industries. The contribution made by this research to the methodological developments in festival and event studies is of additional significance. The application of visual research methods, including semiotic analysis and photo-elicitation within phenomenological interviews, has previously been applied in marketing, consumer, and tourism research, but not to the understanding of festival brands and stakeholders’ brand relationship types. Findings of this research illustrate that existing marketing and consumer brand frameworks and stakeholder theories are applicable to festivals. Further, it is possible to define ‘a’ Fringe brand image which is subjective and contradictory. The unique open-access and organic, operational model of the Fringe facilitates its many contributors, and consumers. Fringe stakeholders may be categorised according to their level of engagement with the Fringe (as primary or secondary) and their particular stakeholder role(s), which are varied and multiple. Fringe-stakeholder brand relationship types are overwhelmingly positive; and are based upon interpersonal relationship dimensions (including friendships, marriages, kinships and partnerships). Fringe-stakeholder brand relationship types can be classified therefore as having similar dimensions to those brand relationship types previously described for consumer products and brands.

Modelling uncertainty in agricultural image analysis

C.M. Onyango, J.A. Marchant and R. Zwiggelaar, 'Modelling uncertainty in agricultural image analysis', Computers and Electronics in Agriculture 17 (3), 295-305 (1997)

Improved Tracking of Multiple Humans with Trajectory Prediction and Occlusion Modeling

A combined 2D, 3D approach is presented that allows for robust tracking of moving bodies in a given environment as observed via a single, uncalibrated video camera. Tracking is robust even in the presence of occlusions. Low-level features are often insufficient for detection, segmentation, and tracking of non-rigid moving objects. Therefore, an improved mechanism is proposed that combines low-level (image processing) and mid-level (recursive trajectory estimation) information obtained during the tracking process. The resulting system can segment and maintain the tracking of moving objects before, during, and after occlusion. At each frame, the system also extracts a stabilized coordinate frame of the moving objects. This stabilized frame is used to resize and resample the moving blob so that it can be used as input to motion recognition modules. The approach enables robust tracking without constraining the system to know the shape of the objects being tracked beforehand; although, some assumptions are made about the characteristics of the shape of the objects, and how they evolve with time. Experiments in tracking moving people are described.

3D Trajectory Recovery for Tracking Multiple Objects and Trajectory Guided Recognition of Actions

A mechanism is proposed that integrates low-level (image processing), mid-level (recursive 3D trajectory estimation), and high-level (action recognition) processes. It is assumed that the system observes multiple moving objects via a single, uncalibrated video camera. A novel extended Kalman filter formulation is used in estimating the relative 3D motion trajectories up to a scale factor. The recursive estimation process provides a prediction and error measure that is exploited in higher-level stages of action recognition. Conversely, higher-level mechanisms provide feedback that allows the system to reliably segment and maintain the tracking of moving objects before, during, and after occlusion. The 3D trajectory, occlusion, and segmentation information are utilized in extracting stabilized views of the moving object. Trajectory-guided recognition (TGR) is proposed as a new and efficient method for adaptive classification of action. The TGR approach is demonstrated using "motion history images" that are then recognized via a mixture of Gaussian classifier. The system was tested in recognizing various dynamic human outdoor activities; e.g., running, walking, roller blading, and cycling. Experiments with synthetic data sets are used to evaluate stability of the trajectory estimator with respect to noise.

Learning Embeddings for Indexing, Retrieval, and Classification, with Applications to Object and Shape Recognition in Image Databases

Nearest neighbor retrieval is the task of identifying, given a database of objects and a query object, the objects in the database that are the most similar to the query. Retrieving nearest neighbors is a necessary component of many practical applications, in fields as diverse as computer vision, pattern recognition, multimedia databases, bioinformatics, and computer networks. At the same time, finding nearest neighbors accurately and efficiently can be challenging, especially when the database contains a large number of objects, and when the underlying distance measure is computationally expensive. This thesis proposes new methods for improving the efficiency and accuracy of nearest neighbor retrieval and classification in spaces with computationally expensive distance measures. The proposed methods are domain-independent, and can be applied in arbitrary spaces, including non-Euclidean and non-metric spaces. In this thesis particular emphasis is given to computer vision applications related to object and shape recognition, where expensive non-Euclidean distance measures are often needed to achieve high accuracy. The first contribution of this thesis is the BoostMap algorithm for embedding arbitrary spaces into a vector space with a computationally efficient distance measure. Using this approach, an approximate set of nearest neighbors can be retrieved efficiently - often orders of magnitude faster than retrieval using the exact distance measure in the original space. The BoostMap algorithm has two key distinguishing features with respect to existing embedding methods. First, embedding construction explicitly maximizes the amount of nearest neighbor information preserved by the embedding. Second, embedding construction is treated as a machine learning problem, in contrast to existing methods that are based on geometric considerations. The second contribution is a method for constructing query-sensitive distance measures for the purposes of nearest neighbor retrieval and classification. In high-dimensional spaces, query-sensitive distance measures allow for automatic selection of the dimensions that are the most informative for each specific query object. It is shown theoretically and experimentally that query-sensitivity increases the modeling power of embeddings, allowing embeddings to capture a larger amount of the nearest neighbor structure of the original space. The third contribution is a method for speeding up nearest neighbor classification by combining multiple embedding-based nearest neighbor classifiers in a cascade. In a cascade, computationally efficient classifiers are used to quickly classify easy cases, and classifiers that are more computationally expensive and also more accurate are only applied to objects that are harder to classify. An interesting property of the proposed cascade method is that, under certain conditions, classification time actually decreases as the size of the database increases, a behavior that is in stark contrast to the behavior of typical nearest neighbor classification systems. The proposed methods are evaluated experimentally in several different applications: hand shape recognition, off-line character recognition, online character recognition, and efficient retrieval of time series. In all datasets, the proposed methods lead to significant improvements in accuracy and efficiency compared to existing state-of-the-art methods. In some datasets, the general-purpose methods introduced in this thesis even outperform domain-specific methods that have been custom-designed for such datasets.

Fast Learning VIEWNET Architectures for Recognizing 3-D Objects from Multiple 2-D Views

The recognition of 3-D objects from sequences of their 2-D views is modeled by a family of self-organizing neural architectures, called VIEWNET, that use View Information Encoded With NETworks. VIEWNET incorporates a preprocessor that generates a compressed but 2-D invariant representation of an image, a supervised incremental learning system that classifies the preprocessed representations into 2-D view categories whose outputs arc combined into 3-D invariant object categories, and a working memory that makes a 3-D object prediction by accumulating evidence from 3-D object category nodes as multiple 2-D views are experienced. The simplest VIEWNET achieves high recognition scores without the need to explicitly code the temporal order of 2-D views in working memory. Working memories are also discussed that save memory resources by implicitly coding temporal order in terms of the relative activity of 2-D view category nodes, rather than as explicit 2-D view transitions. Variants of the VIEWNET architecture may also be used for scene understanding by using a preprocessor and classifier that can determine both What objects are in a scene and Where they are located. The present VIEWNET preprocessor includes the CORT-X 2 filter, which discounts the illuminant, regularizes and completes figural boundaries, and suppresses image noise. This boundary segmentation is rendered invariant under 2-D translation, rotation, and dilation by use of a log-polar transform. The invariant spectra undergo Gaussian coarse coding to further reduce noise and 3-D foreshortening effects, and to increase generalization. These compressed codes are input into the classifier, a supervised learning system based on the fuzzy ARTMAP algorithm. Fuzzy ARTMAP learns 2-D view categories that are invariant under 2-D image translation, rotation, and dilation as well as 3-D image transformations that do not cause a predictive error. Evidence from sequence of 2-D view categories converges at 3-D object nodes that generate a response invariant under changes of 2-D view. These 3-D object nodes input to a working memory that accumulates evidence over time to improve object recognition. ln the simplest working memory, each occurrence (nonoccurrence) of a 2-D view category increases (decreases) the corresponding node's activity in working memory. The maximally active node is used to predict the 3-D object. Recognition is studied with noisy and clean image using slow and fast learning. Slow learning at the fuzzy ARTMAP map field is adapted to learn the conditional probability of the 3-D object given the selected 2-D view category. VIEWNET is demonstrated on an MIT Lincoln Laboratory database of l28x128 2-D views of aircraft with and without additive noise. A recognition rate of up to 90% is achieved with one 2-D view and of up to 98.5% correct with three 2-D views. The properties of 2-D view and 3-D object category nodes are compared with those of cells in monkey inferotemporal cortex.

A Neural Model of 3D Shape-From-Texture: Multiple-Scale Filtering, Boundary Grouping, and Surface Filling-In

A neural model is presented of how cortical areas V1, V2, and V4 interact to convert a textured 2D image into a representation of curved 3D shape. Two basic problems are solved to achieve this: (1) Patterns of spatially discrete 2D texture elements are transformed into a spatially smooth surface representation of 3D shape. (2) Changes in the statistical properties of texture elements across space induce the perceived 3D shape of this surface representation. This is achieved in the model through multiple-scale filtering of a 2D image, followed by a cooperative-competitive grouping network that coherently binds texture elements into boundary webs at the appropriate depths using a scale-to-depth map and a subsequent depth competition stage. These boundary webs then gate filling-in of surface lightness signals in order to form a smooth 3D surface percept. The model quantitatively simulates challenging psychophysical data about perception of prolate ellipsoids (Todd and Akerstrom, 1987, J. Exp. Psych., 13, 242). In particular, the model represents a high degree of 3D curvature for a certain class of images, all of whose texture elements have the same degree of optical compression, in accordance with percepts of human observers. Simulations of 3D percepts of an elliptical cylinder, a slanted plane, and a photo of a golf ball are also presented.

Intermolecular Multiple Quantum Coherences Enable Accurate Thermal Imaging of Red Bone Marrow During Thermal Therapy of Bone Metastases

Prostate and breast cancers are two of the most common types of cancer in the United States, and those cancers metastasize to bone in more than two thirds of patients. Recent evidence suggests that thermal therapy is effective at treating metastatic bone cancer. For example, thermal therapy enables targeted drug delivery to bone, ablation of cancer cells in bone marrow, and palliation of bone pain. Thermal therapy of bone metastases would be greatly improved if it were possible to image the temperature of the tissue surrounding the disease, which is usually red bone marrow (RBM). Unfortunately, current thermal imaging techniques are inaccurate in RBM.

This dissertation shows that many of the difficulties with thermal imaging of RBM can be overcome using a magnetic resonance phenomenon called an intermolecular multiple quantum coherence (iMQC). Herein, iMQCs are detected with a magnetic resonance imaging (MRI) pulse sequence called multi-spin-echo HOMOGENIZED with off resonance transfer (MSE-HOT). Compared to traditional methods, MSE-HOT provided ten-fold more accurate images of temperature change. Furthermore, MSE-HOT was translated to a human MRI scanner, which enabled imaging of RBM temperature during heating with a clinical focused ultrasound applicator. In summary, this dissertation develops a MRI technique that enables thermal imaging of RBM during thermal therapy of bone metastases.

Time-reversal ultrawideband breast imaging: pulse design criteria considering multiple tumors with unknown tissue properties

Pulse design is investigated for time-reversal (TR) imaging as applied to ultrawideband (UWB) breast cancer detection. Earlier it has been shown that a suitably-designed UWB pulse may help to improve imaging performance for a single-tumor breast phantom with predetermined lesion properties. The current work considers the following more general and practical situations: presence of multiple malignancies with unknown tumor size and dielectric properties. Four pulse selection criteria are proposed with each focusing on one of the following aspects: eliminating signal clutter generated by tissue inhomogeneities, canceling mutual interference among tumors, improving image resolution, and suppressing artifacts created by sidelobe of the target response. By applying the proposed criteria, the shape parameters of UWB waveforms with desirable characteristics are identified through search of all the possible pulses. Simulation example using a numerical breast phantom, comprised of two tumors and structured clutter distribution, demonstrates the effectiveness of the proposed approach. Specifically, a tradeoff between the image resolution and signal-to-clutter contrast (SCC) is observed in terms of selection of the excitation waveforms.

Tight junctional abnormality in multiple sclerosis white matter affects all calibres of vessel and is associated with blood-brain barrier leakage and active demyelination

Blood-brain barrier (BBB) hyperpermeability in multiple sclerosis (MS) is associated with lesion pathogenesis and has been linked to pathology in microvascular tight junctions (TJs). This study quantifies the uneven distribution of TJ pathology and its association with BBB leakage. Frozen sections from plaque and normal-appearing white matter (NAWM) in 14 cases were studied together with white matter from six neurological and five normal controls. Using single and double immunofluorescence and confocal microscopy, the TJ-associated protein zonula occludens-1 (ZO-1) was examined across lesion types and tissue categories, and in relation to fibrinogen leakage. Confocal image data sets were analysed for 2198 MS and 1062 control vessels. Significant differences in the incidence of TJ abnormalities were detected between the different lesion types in MS and between MS and control white matter. These were frequent in oil-red O (ORO)+ active plaques, affecting 42% of vessel segments, but less frequent in ORO- inactive plaques (23%), NAWM (13%), and normal (3.7%) and neurological controls (8%). A similar pattern was found irrespective of the vessel size, supporting a causal role for diffusible inflammatory mediators. In both NAWM and inactive lesions, dual labelling showed that vessels with the most TJ abnormality also showed most fibrinogen leakage. This was even more pronounced in active lesions, where 41% of vessels in the highest grade for TJ alteration showed severe leakage. It is concluded that disruption of TJs in MS, affecting both paracellular and transcellular paths, contributes to BBB leakage. TJ abnormality and BBB leakage in inactive lesions suggests either failure of TJ repair or a continuing pathological process. In NAWM, it suggests either pre-lesional change or secondary damage. Clinically inapparent TJ pathology has prognostic implications and should be considered when planning disease-modifying therapy

Hybrid image/ray-shooting UHF radio propagation predictor for populated indoor environments

A novel model for indoor wireless communication, based on a dual image and ray-shooting approach, is presented. The model, capable of improved site-specific indoor propagation prediction, considers multiple human bodies moving within the environment. In a modern office at 2.45GHz, the combined effect of pedestrian traffic and a moving receiver causes rapid temporal fading of up to 30dB.

A Robust Co-Localisation Measurement Utilising Z-Stack Image Intensity Similarities for Biological Studies

Background: Co-localisation is a widely used measurement in immunohistochemical analysis to determine if fluorescently labelled biological entities, such as cells, proteins or molecules share a same location. However the measurement of co-localisation is challenging due to the complex nature of such fluorescent images, especially when multiple focal planes are captured. The current state-of-art co-localisation measurements of 3-dimensional (3D) image stacks are biased by noise and cross-overs from non-consecutive planes.

Method: In this study, we have developed Co-localisation Intensity Coefficients (CICs) and Co-localisation Binary Coefficients (CBCs), which uses rich z-stack data from neighbouring focal planes to identify similarities between image intensities of two and potentially more fluorescently-labelled biological entities. This was developed using z-stack images from murine organotypic slice cultures from central nervous system tissue, and two sets of pseudo-data. A large amount of non-specific cross-over situations are excluded using this method. This proposed method is also proven to be robust in recognising co-localisations even when images are polluted with a range of noises.

Results: The proposed CBCs and CICs produce robust co-localisation measurements which are easy to interpret, resilient to noise and capable of removing a large amount of false positivity, such as non-specific cross-overs. Performance of this method of measurement is significantly more accurate than existing measurements, as determined statistically using pseudo datasets of known values. This method provides an important and reliable tool for fluorescent 3D neurobiological studies, and will benefit other biological studies which measure fluorescence co-localisation in 3D.

AdaBoost Multiple Feature Selection and Combination for Face Recognition

Gabor features have been recognized as one of the most successful face representations. Encouraged by the results given by this approach, other kind of facial representations based on Steerable Gaussian first order kernels and Harris corner detector are proposed in this paper. In order to reduce the high dimensional feature space, PCA and LDA techniques are employed. Once the features have been extracted, AdaBoost learning algorithm is used to select and combine the most representative features. The experimental results on XM2VTS database show an encouraging recognition rate, showing an important improvement with respect to face descriptors only based on Gabor filters.