Computer System for Visual Recognition Using Active Knowledge

A system for visual recognition is described, with implications for the general problem of representation of knowledge to assist control. The immediate objective is a computer system that will recognize objects in a visual scene, specifically hammers. The computer receives an array of light intensities from a device like a television camera. It is to locate and identify the hammer if one is present. The computer must produce from the numerical "sensory data" a symbolic description that constitutes its perception of the scene. Of primary concern is the control of the recognition process. Control decisions should be guided by the partial results obtained on the scene. If a hammer handle is observed this should suggest that the handle is part of a hammer and advise where to look for the hammer head. The particular knowledge that a handle has been found combines with general knowledge about hammers to influence the recognition process. This use of knowledge to direct control is denoted here by the term "active knowledge". A descriptive formalism is presented for visual knowledge which identifies the relationships relevant to the active use of the knowledge. A control structure is provided which can apply knowledge organized in this fashion actively to the processing of a given scene.

Biologically Inspired Approaches to Automated Feature Extraction and Target Recognition

Ongoing research at Boston University has produced computational models of biological vision and learning that embody a growing corpus of scientific data and predictions. Vision models perform long-range grouping and figure/ground segmentation, and memory models create attentionally controlled recognition codes that intrinsically cornbine botton-up activation and top-down learned expectations. These two streams of research form the foundation of novel dynamically integrated systems for image understanding. Simulations using multispectral images illustrate road completion across occlusions in a cluttered scene and information fusion from incorrect labels that are simultaneously inconsistent and correct. The CNS Vision and Technology Labs (cns.bu.edulvisionlab and cns.bu.edu/techlab) are further integrating science and technology through analysis, testing, and development of cognitive and neural models for large-scale applications, complemented by software specification and code distribution.

An Active Pattern Recognition Architecture for Mobile Robots

An active, attentionally-modulated recognition architecture is proposed for object recognition and scene analysis. The proposed architecture forms part of navigation and trajectory planning modules for mobile robots. Key characteristics of the system include movement planning and execution based on environmental factors and internal goal definitions. Real-time implementation of the system is based on space-variant representation of the visual field, as well as an optimal visual processing scheme utilizing separate and parallel channels for the extraction of boundaries and stimulus qualities. A spatial and temporal grouping module (VWM) allows for scene scanning, multi-object segmentation, and featural/object priming. VWM is used to modulate a tn~ectory formation module capable of redirecting the focus of spatial attention. Finally, an object recognition module based on adaptive resonance theory is interfaced through VWM to the visual processing module. The system is capable of using information from different modalities to disambiguate sensory input.

The What-And-Where Filter: A Spatial Mapping Neural Network for Object Recognition and Image Understanding

The What-and-Where filter forms part of a neural network architecture for spatial mapping, object recognition, and image understanding. The Where fllter responds to an image figure that has been separated from its background. It generates a spatial map whose cell activations simultaneously represent the position, orientation, ancl size of all tbe figures in a scene (where they are). This spatial map may he used to direct spatially localized attention to these image features. A multiscale array of oriented detectors, followed by competitve and interpolative interactions between position, orientation, and size scales, is used to define the Where filter. This analysis discloses several issues that need to be dealt with by a spatial mapping system that is based upon oriented filters, such as the role of cliff filters with and without normalization, the double peak problem of maximum orientation across size scale, and the different self-similar interpolation properties across orientation than across size scale. Several computationally efficient Where filters are proposed. The Where filter rnay be used for parallel transformation of multiple image figures into invariant representations that are insensitive to the figures' original position, orientation, and size. These invariant figural representations form part of a system devoted to attentive object learning and recognition (what it is). Unlike some alternative models where serial search for a target occurs, a What and Where representation can he used to rapidly search in parallel for a desired target in a scene. Such a representation can also be used to learn multidimensional representations of objects and their spatial relationships for purposes of image understanding. The What-and-Where filter is inspired by neurobiological data showing that a Where processing stream in the cerebral cortex is used for attentive spatial localization and orientation, whereas a What processing stream is used for attentive object learning and recognition.

Neural mechanisms of context effects on face recognition: automatic binding and context shift decrements.

Although people do not normally try to remember associations between faces and physical contexts, these associations are established automatically, as indicated by the difficulty of recognizing familiar faces in different contexts ("butcher-on-the-bus" phenomenon). The present fMRI study investigated the automatic binding of faces and scenes. In the face-face (F-F) condition, faces were presented alone during both encoding and retrieval, whereas in the face/scene-face (FS-F) condition, they were presented overlaid on scenes during encoding but alone during retrieval (context change). Although participants were instructed to focus only on the faces during both encoding and retrieval, recognition performance was worse in the FS-F than in the F-F condition ("context shift decrement" [CSD]), confirming automatic face-scene binding during encoding. This binding was mediated by the hippocampus as indicated by greater subsequent memory effects (remembered > forgotten) in this region for the FS-F than the F-F condition. Scene memory was mediated by right parahippocampal cortex, which was reactivated during successful retrieval when the faces were associated with a scene during encoding (FS-F condition). Analyses using the CSD as a regressor yielded a clear hemispheric asymmetry in medial temporal lobe activity during encoding: Left hippocampal and parahippocampal activity was associated with a smaller CSD, indicating more flexible memory representations immune to context changes, whereas right hippocampal/rhinal activity was associated with a larger CSD, indicating less flexible representations sensitive to context change. Taken together, the results clarify the neural mechanisms of context effects on face recognition.

A case study in identifying acceptable bitrates for human face recognition tasks

Face recognition from images or video footage requires a certain level of recorded image quality. This paper derives acceptable bitrates (relating to levels of compression and consequently quality) of footage with human faces, using an industry implementation of the standard H.264/MPEG-4 AVC and the Closed-Circuit Television (CCTV) recording systems on London buses. The London buses application is utilized as a case study for setting up a methodology and implementing suitable data analysis for face recognition from recorded footage, which has been degraded by compression. The majority of CCTV recorders on buses use a proprietary format based on the H.264/MPEG-4 AVC video coding standard, exploiting both spatial and temporal redundancy. Low bitrates are favored in the CCTV industry for saving storage and transmission bandwidth, but they compromise the image usefulness of the recorded imagery. In this context, usefulness is determined by the presence of enough facial information remaining in the compressed image to allow a specialist to recognize a person. The investigation includes four steps: (1) Development of a video dataset representative of typical CCTV bus scenarios. (2) Selection and grouping of video scenes based on local (facial) and global (entire scene) content properties. (3) Psychophysical investigations to identify the key scenes, which are most affected by compression, using an industry implementation of H.264/MPEG-4 AVC. (4) Testing of CCTV recording systems on buses with the key scenes and further psychophysical investigations. The results showed a dependency upon scene content properties. Very dark scenes and scenes with high levels of spatial–temporal busyness were the most challenging to compress, requiring higher bitrates to maintain useful information.

Attentional Selection in Object Recognition

A key problem in object recognition is selection, namely, the problem of identifying regions in an image within which to start the recognition process, ideally by isolating regions that are likely to come from a single object. Such a selection mechanism has been found to be crucial in reducing the combinatorial search involved in the matching stage of object recognition. Even though selection is of help in recognition, it has largely remained unsolved because of the difficulty in isolating regions belonging to objects under complex imaging conditions involving occlusions, changing illumination, and object appearances. This thesis presents a novel approach to the selection problem by proposing a computational model of visual attentional selection as a paradigm for selection in recognition. In particular, it proposes two modes of attentional selection, namely, attracted and pay attention modes as being appropriate for data and model-driven selection in recognition. An implementation of this model has led to new ways of extracting color, texture and line group information in images, and their subsequent use in isolating areas of the scene likely to contain the model object. Among the specific results in this thesis are: a method of specifying color by perceptual color categories for fast color region segmentation and color-based localization of objects, and a result showing that the recognition of texture patterns on model objects is possible under changes in orientation and occlusions without detailed segmentation. The thesis also presents an evaluation of the proposed model by integrating with a 3D from 2D object recognition system and recording the improvement in performance. These results indicate that attentional selection can significantly overcome the computational bottleneck in object recognition, both due to a reduction in the number of features, and due to a reduction in the number of matches during recognition using the information derived during selection. Finally, these studies have revealed a surprising use of selection, namely, in the partial solution of the pose of a 3D object.

Using appearance and context for outdoor scene object classification

We propose a probabilistic object classifier for outdoor scene analysis as a first step in solving the problem of scene context generation. The method begins with a top-down control, which uses the previously learned models (appearance and absolute location) to obtain an initial pixel-level classification. This information provides us the core of objects, which is used to acquire a more accurate object model. Therefore, their growing by specific active regions allows us to obtain an accurate recognition of known regions. Next, a stage of general segmentation provides the segmentation of unknown regions by a bottom-strategy. Finally, the last stage tries to perform a region fusion of known and unknown segmented objects. The result is both a segmentation of the image and a recognition of each segment as a given object class or as an unknown segmented object. Furthermore, experimental results are shown and evaluated to prove the validity of our proposal

Object and Scene Classification: what does a Supervised Approach Provide us?

Given a set of images of scenes containing different object categories (e.g. grass, roads) our objective is to discover these objects in each image, and to use this object occurrences to perform a scene classification (e.g. beach scene, mountain scene). We achieve this by using a supervised learning algorithm able to learn with few images to facilitate the user task. We use a probabilistic model to recognise the objects and further we classify the scene based on their object occurrences. Experimental results are shown and evaluated to prove the validity of our proposal. Object recognition performance is compared to the approaches of He et al. (2004) and Marti et al. (2001) using their own datasets. Furthermore an unsupervised method is implemented in order to evaluate the advantages and disadvantages of our supervised classification approach versus an unsupervised one

Scene Classification Using a Hybrid Generative/Discriminative Approach

We investigate whether dimensionality reduction using a latent generative model is beneficial for the task of weakly supervised scene classification. In detail, we are given a set of labeled images of scenes (for example, coast, forest, city, river, etc.), and our objective is to classify a new image into one of these categories. Our approach consists of first discovering latent ";topics"; using probabilistic Latent Semantic Analysis (pLSA), a generative model from the statistical text literature here applied to a bag of visual words representation for each image, and subsequently, training a multiway classifier on the topic distribution vector for each image. We compare this approach to that of representing each image by a bag of visual words vector directly and training a multiway classifier on these vectors. To this end, we introduce a novel vocabulary using dense color SIFT descriptors and then investigate the classification performance under changes in the size of the visual vocabulary, the number of latent topics learned, and the type of discriminative classifier used (k-nearest neighbor or SVM). We achieve superior classification performance to recent publications that have used a bag of visual word representation, in all cases, using the authors' own data sets and testing protocols. We also investigate the gain in adding spatial information. We show applications to image retrieval with relevance feedback and to scene classification in videos

Use of decision trees in colour feature selection: application to object recognition in outdoor scenes

A new method for the automated selection of colour features is described. The algorithm consists of two stages of processing. In the first, a complete set of colour features is calculated for every object of interest in an image. In the second stage, each object is mapped into several n-dimensional feature spaces in order to select the feature set with the smallest variables able to discriminate the remaining objects. The evaluation of the discrimination power for each concrete subset of features is performed by means of decision trees composed of linear discrimination functions. This method can provide valuable help in outdoor scene analysis where no colour space has been demonstrated as being the most suitable. Experiment results recognizing objects in outdoor scenes are reported

Model-based objects recognition in industrial environments for autonomous vehicles control

Behavior-based navigation of autonomous vehicles requires the recognition of the navigable areas and the potential obstacles. In this paper we describe a model-based objects recognition system which is part of an image interpretation system intended to assist the navigation of autonomous vehicles that operate in industrial environments. The recognition system integrates color, shape and texture information together with the location of the vanishing point. The recognition process starts from some prior scene knowledge, that is, a generic model of the expected scene and the potential objects. The recognition system constitutes an approach where different low-level vision techniques extract a multitude of image descriptors which are then analyzed using a rule-based reasoning system to interpret the image content. This system has been implemented using a rule-based cooperative expert system

Model-based objects recognition in man-made environments

We describe a model-based objects recognition system which is part of an image interpretation system intended to assist autonomous vehicles navigation. The system is intended to operate in man-made environments. Behavior-based navigation of autonomous vehicles involves the recognition of navigable areas and the potential obstacles. The recognition system integrates color, shape and texture information together with the location of the vanishing point. The recognition process starts from some prior scene knowledge, that is, a generic model of the expected scene and the potential objects. The recognition system constitutes an approach where different low-level vision techniques extract a multitude of image descriptors which are then analyzed using a rule-based reasoning system to interpret the image content. This system has been implemented using CEES, the C++ embedded expert system shell developed in the Systems Engineering and Automatic Control Laboratory (University of Girona) as a specific rule-based problem solving tool. It has been especially conceived for supporting cooperative expert systems, and uses the object oriented programming paradigm

Visual Object Recognition Using Deformable Models of Vehicles

This workshop paper reports recent developments to a vision system for traffic interpretation which relies extensively on the use of geometrical and scene context. Firstly, a new approach to pose refinement is reported, based on forces derived from prominent image derivatives found close to an initial hypothesis. Secondly, a parameterised vehicle model is reported, able to represent different vehicle classes. This general vehicle model has been fitted to sample data, and subjected to a Principal Component Analysis to create a deformable model of common car types having 6 parameters. We show that the new pose recovery technique is also able to operate on the PCA model, to allow the structure of an initial vehicle hypothesis to be adapted to fit the prevailing context. We report initial experiments with the model, which demonstrate significant improvements to pose recovery.

A novel shape feature for fast region-based pedestrian recognition

A new class of shape features for region classification and high-level recognition is introduced. The novel Randomised Region Ray (RRR) features can be used to train binary decision trees for object category classification using an abstract representation of the scene. In particular we address the problem of human detection using an over segmented input image. We therefore do not rely on pixel values for training, instead we design and train specialised classifiers on the sparse set of semantic regions which compose the image. Thanks to the abstract nature of the input, the trained classifier has the potential to be fast and applicable to extreme imagery conditions. We demonstrate and evaluate its performance in people detection using a pedestrian dataset.