A unified approach to the recognition of complex actions from sequences of zone-crossings

We present a method for the recognition of complex actions. Our method combines automatic learning of simple actions and manual definition of complex actions in a single grammar. Contrary to the general trend in complex action recognition that consists in dividing recognition into two stages, our method performs recognition of simple and complex actions in a unified way. This is performed by encoding simple action HMMs within the stochastic grammar that models complex actions. This unified approach enables a more effective influence of the higher activity layers into the recognition of simple actions which leads to a substantial improvement in the classification of complex actions. We consider the recognition of complex actions based on person transits between areas in the scene. As input, our method receives crossings of tracks along a set of zones which are derived using unsupervised learning of the movement patterns of the objects in the scene. We evaluate our method on a large dataset showing normal, suspicious and threat behaviour on a parking lot. Experiments show an improvement of ~ 30% in the recognition of both high-level scenarios and their composing simple actions with respect to a two-stage approach. Experiments with synthetic noise simulating the most common tracking failures show that our method only experiences a limited decrease in performance when moderate amounts of noise are added.

Image recognition: Visual grouping, recognition, and learning

Vision extracts useful information from images. Reconstructing the three-dimensional structure of our environment and recognizing the objects that populate it are among the most important functions of our visual system. Computer vision researchers study the computational principles of vision and aim at designing algorithms that reproduce these functions. Vision is difficult: the same scene may give rise to very different images depending on illumination and viewpoint. Typically, an astronomical number of hypotheses exist that in principle have to be analyzed to infer a correct scene description. Moreover, image information might be extracted at different levels of spatial and logical resolution dependent on the image processing task. Knowledge of the world allows the visual system to limit the amount of ambiguity and to greatly simplify visual computations. We discuss how simple properties of the world are captured by the Gestalt rules of grouping, how the visual system may learn and organize models of objects for recognition, and how one may control the complexity of the description that the visual system computes.

Towards a real-time 3D object recognition pipeline on mobile GPGPU computing platforms using low-cost RGB-D sensors

In this project, we propose the implementation of a 3D object recognition system which will be optimized to operate under demanding time constraints. The system must be robust so that objects can be recognized properly in poor light conditions and cluttered scenes with significant levels of occlusion. An important requirement must be met: the system must exhibit a reasonable performance running on a low power consumption mobile GPU computing platform (NVIDIA Jetson TK1) so that it can be integrated in mobile robotics systems, ambient intelligence or ambient assisted living applications. The acquisition system is based on the use of color and depth (RGB-D) data streams provided by low-cost 3D sensors like Microsoft Kinect or PrimeSense Carmine. The range of algorithms and applications to be implemented and integrated will be quite broad, ranging from the acquisition, outlier removal or filtering of the input data and the segmentation or characterization of regions of interest in the scene to the very object recognition and pose estimation. Furthermore, in order to validate the proposed system, we will create a 3D object dataset. It will be composed by a set of 3D models, reconstructed from common household objects, as well as a handful of test scenes in which those objects appear. The scenes will be characterized by different levels of occlusion, diverse distances from the elements to the sensor and variations on the pose of the target objects. The creation of this dataset implies the additional development of 3D data acquisition and 3D object reconstruction applications. The resulting system has many possible applications, ranging from mobile robot navigation and semantic scene labeling to human-computer interaction (HCI) systems based on visual information.

A Novel Prediction Method for Early Recognition of Global Human Behaviour in Image Sequences

Human behaviour recognition has been, and still remains, a challenging problem that involves different areas of computational intelligence. The automated understanding of people activities from video sequences is an open research topic in which the computer vision and pattern recognition areas have made big efforts. In this paper, the problem is studied from a prediction point of view. We propose a novel method able to early detect behaviour using a small portion of the input, in addition to the capabilities of it to predict behaviour from new inputs. Specifically, we propose a predictive method based on a simple representation of trajectories of a person in the scene which allows a high level understanding of the global human behaviour. The representation of the trajectory is used as a descriptor of the activity of the individual. The descriptors are used as a cue of a classification stage for pattern recognition purposes. Classifiers are trained using the trajectory representation of the complete sequence. However, partial sequences are processed to evaluate the early prediction capabilities having a specific observation time of the scene. The experiments have been carried out using the three different dataset of the CAVIAR database taken into account the behaviour of an individual. Additionally, different classic classifiers have been used for experimentation in order to evaluate the robustness of the proposal. Results confirm the high accuracy of the proposal on the early recognition of people behaviours.

Visual perception for the 3D recognition of geometric pieces in robotic manipulation

During grasping and intelligent robotic manipulation tasks, the camera position relative to the scene changes dramatically because the robot is moving to adapt its path and correctly grasp objects. This is because the camera is mounted at the robot effector. For this reason, in this type of environment, a visual recognition system must be implemented to recognize and “automatically and autonomously” obtain the positions of objects in the scene. Furthermore, in industrial environments, all objects that are manipulated by robots are made of the same material and cannot be differentiated by features such as texture or color. In this work, first, a study and analysis of 3D recognition descriptors has been completed for application in these environments. Second, a visual recognition system designed from specific distributed client-server architecture has been proposed to be applied in the recognition process of industrial objects without these appearance features. Our system has been implemented to overcome problems of recognition when the objects can only be recognized by geometric shape and the simplicity of shapes could create ambiguity. Finally, some real tests are performed and illustrated to verify the satisfactory performance of the proposed system.

Methodological aspects of scene segmentation.

"COO-2118-0028."

A stable graph-based representation for object recognition through high-order matching

Many Object recognition techniques perform some flavour of point pattern matching between a model and a scene. Such points are usually selected through a feature detection algorithm that is robust to a class of image transformations and a suitable descriptor is computed over them in order to get a reliable matching. Moreover, some approaches take an additional step by casting the correspondence problem into a matching between graphs defined over feature points. The motivation is that the relational model would add more discriminative power, however the overall effectiveness strongly depends on the ability to build a graph that is stable with respect to both changes in the object appearance and spatial distribution of interest points. In fact, widely used graph-based representations, have shown to suffer some limitations, especially with respect to changes in the Euclidean organization of the feature points. In this paper we introduce a technique to build relational structures over corner points that does not depend on the spatial distribution of the features. © 2012 ICPR Org Committee.

Color-based surface reflectance separation for scene illumination estimation and rendering

Given the importance of color processing in computer vision and computer graphics, estimating and rendering illumination spectral reflectance of image scenes is important to advance the capability of a large class of applications such as scene reconstruction, rendering, surface segmentation, object recognition, and reflectance estimation. Consequently, this dissertation proposes effective methods for reflection components separation and rendering in single scene images. Based on the dichromatic reflectance model, a novel decomposition technique, named the Mean-Shift Decomposition (MSD) method, is introduced to separate the specular from diffuse reflectance components. This technique provides a direct access to surface shape information through diffuse shading pixel isolation. More importantly, this process does not require any local color segmentation process, which differs from the traditional methods that operate by aggregating color information along each image plane. ^ Exploiting the merits of the MSD method, a scene illumination rendering technique is designed to estimate the relative contributing specular reflectance attributes of a scene image. The image feature subset targeted provides a direct access to the surface illumination information, while a newly introduced efficient rendering method reshapes the dynamic range distribution of the specular reflectance components over each image color channel. This image enhancement technique renders the scene illumination reflection effectively without altering the scene’s surface diffuse attributes contributing to realistic rendering effects. ^ As an ancillary contribution, an effective color constancy algorithm based on the dichromatic reflectance model was also developed. This algorithm selects image highlights in order to extract the prominent surface reflectance that reproduces the exact illumination chromaticity. This evaluation is presented using a novel voting scheme technique based on histogram analysis. ^ In each of the three main contributions, empirical evaluations were performed on synthetic and real-world image scenes taken from three different color image datasets. The experimental results show over 90% accuracy in illumination estimation contributing to near real world illumination rendering effects. ^

Contributions Of the Human Medial Prefrontal Cortex To Associative Recognition Memory: Evidence From Functional Neuroimaging

Neuroimaging studies of episodic memory, or memory of events from our personal past, have predominantly focused their attention on medial temporal lobe (MTL). There is growing acknowledgement however, from the cognitive neuroscience of memory literature, that regions outside the MTL can support episodic memory processes. The medial prefrontal cortex is one such region garnering increasing interest from researchers. Using behavioral and functional magnetic resonance imaging measures, over two studies, this thesis provides evidence of a mnemonic role of the medial PFC. In the first study, participants were scanned while judging the extent to which they agreed or disagreed with the sociopolitical views of unfamiliar individuals. Behavioral tests of associative recognition revealed that participants remembered with high confidence viewpoints previously linked with judgments of strong agreement/disagreement. Neurally, the medial PFC mediated the interaction between high-confidence associative recognition memory and beliefs associated with strong agree/disagree judgments. In an effort to generalize this finding to well-established associative information, in the second study, we investigated associative recognition memory for real-world concepts. Object-scene pairs congruent or incongruent with a preexisting schema were presented to participants in a cued-recall paradigm. Behavioral tests of conceptual and perceptual recognition revealed memory enhancements arising from strong resonance between presented pairs and preexisting schemas. Neurally, the medial PFC tracked increases in visual recall of schema-congruent pairs whereas the MTL tracked increases in visual recall of schema-incongruent pairs. Additionally, ventral areas of the medial PFC tracked conceptual components of visual recall specifically for schema-congruent pairs. These findings are consistent with a recent theoretical proposal of medial PFC contributions to memory for schema-related content. Collectively, these studies provide evidence of a role for the medial PFC in associative recognition memory persisting for associative information deployed in our daily social interactions and for those associations formed over multiple learning episodes. Additionally, this set of findings advance our understanding of the cognitive contributions of the medial PFC beyond its canonical role in processes underlying social cognition.

Image usefulness of compressed surveillance footage with different scene contents

The police use both subjective (i.e. police staff) and automated (e.g. face recognition systems) methods for the completion of visual tasks (e.g person identification). Image quality for police tasks has been defined as the image usefulness, or image suitability of the visual material to satisfy a visual task. It is not necessarily affected by any artefact that may affect the visual image quality (i.e. decrease fidelity), as long as these artefacts do not affect the relevant useful information for the task. The capture of useful information will be affected by the unconstrained conditions commonly encountered by CCTV systems such as variations in illumination and high compression levels. The main aim of this thesis is to investigate aspects of image quality and video compression that may affect the completion of police visual tasks/applications with respect to CCTV imagery. This is accomplished by investigating 3 specific police areas/tasks utilising: 1) the human visual system (HVS) for a face recognition task, 2) automated face recognition systems, and 3) automated human detection systems. These systems (HVS and automated) were assessed with defined scene content properties, and video compression, i.e. H.264/MPEG-4 AVC. The performance of imaging systems/processes (e.g. subjective investigations, performance of compression algorithms) are affected by scene content properties. No other investigation has been identified that takes into consideration scene content properties to the same extend. Results have shown that the HVS is more sensitive to compression effects in comparison to the automated systems. In automated face recognition systems, `mixed lightness' scenes were the most affected and `low lightness' scenes were the least affected by compression. In contrast the HVS for the face recognition task, `low lightness' scenes were the most affected and `medium lightness' scenes the least affected. For the automated human detection systems, `close distance' and `run approach' are some of the most commonly affected scenes. Findings have the potential to broaden the methods used for testing imaging systems for security applications.

Robust indoor speaker recognition in a network of audio and video sensors

Situational awareness is achieved naturally by the human senses of sight and hearing in combination. Automatic scene understanding aims at replicating this human ability using microphones and cameras in cooperation. In this paper, audio and video signals are fused and integrated at different levels of semantic abstractions. We detect and track a speaker who is relatively unconstrained, i.e., free to move indoors within an area larger than the comparable reported work, which is usually limited to round table meetings. The system is relatively simple: consisting of just 4 microphone pairs and a single camera. Results show that the overall multimodal tracker is more reliable than single modality systems, tolerating large occlusions and cross-talk. System evaluation is performed on both single and multi-modality tracking. The performance improvement given by the audio–video integration and fusion is quantified in terms of tracking precision and accuracy as well as speaker diarisation error rate and precision–recall (recognition). Improvements vs. the closest works are evaluated: 56% sound source localisation computational cost over an audio only system, 8% speaker diarisation error rate over an audio only speaker recognition unit and 36% on the precision–recall metric over an audio–video dominant speaker recognition method.

Exploring the effectiveness of virtual product placement on recall and recognition of new brands

This paper investigates the effectiveness of virtual product placement as a marketing tool by examining the relationship between brand recall and recognition and virtual product placement. It also aims to address a gap in the existing academic literature by focusing on the impact of product placement on recall and recognition of new brands. The growing importance of product placement is discussed and a review of previous research on product placement and virtual product placement is provided. The research methodology used to study the recall and recognition effects of virtual product placement are described and key findings presented. Finally, implications are discussed and recommendations for future research provided.