Dimitrios Makris

Learning semantic scene models from observing activity in visual surveillance

This paper considers the problem of automatically learning an activity-based semantic scene model from a stream of video data. A scene model is proposed that labels regions according to an identifiable activity in each region, such as entry/exit zones, junctions, paths, and stop zones. We present several unsupervised methods that learn these scene elements and present results that show the efficiency of our approach. Finally, we describe how the models can be used to support the interpretation of moving objects in a visual surveillance environment.

Fall detection system using Kinect's infrared sensor

This paper presents a novel fall detection system based on the Kinect sensor. The system runs in real-time and is capable of detecting walking falls accurately and robustly without taking into account any false positive activities (i.e. lying on the floor). Velocity and inactivity calculations are performed to decide whether a fall has occurred. The key novelty of our approach is measuring the velocity based on the contraction or expansion of the width, height and depth of the 3D bounding box. By explicitly using the 3D bounding box, our algorithm requires no pre-knowledge of the scene (i.e. floor), as the set of detected actions are adequate to complete the process of fall detection.

Path detection in video surveillance

Abstract This paper addresses the problem of automatically extracting frequently used pedestrian pathways from video sequences of natural outdoor scenes. Path models are learnt from the accumulation of trajectory data over long time periods, and can be used to augment the classification of subsequent track data. In particular, labelled paths provide an efficient means for compressing the trajectory data for logging purposes. In addition, the model can be used to compute a probabilistic prediction of the pedestrians location many time steps ahead, and to aid the recognition of unusual behaviour identified as atypical object motion.

Automatic learning of an activity-based semantic scene model

The paper proposes an activity-based semantic model for a scene under visual surveillance. It illustrates methods that allow unsupervised learning of the model from trajectory data derived from automatic visual surveillance cameras. Results are shown for each method. Finally, the benefits of such a model in a visual surveillance system are discussed.

G3D: A gaming action dataset and real time action recognition evaluation framework

In this paper a novel evaluation framework for measuring the performance of real-time action recognition methods is presented. The evaluation framework will extend the time-based event detection metric to model multiple distinct action classes. The proposed metric provides more accurate indications of the performance of action recognition algorithms for games and other similar applications since it takes into consideration restrictions related to time and consecutive repetitions. Furthermore, a new dataset, G3D for real-time action recognition in gaming containing synchronised video, depth and skeleton data is provided. Our results indicate the need of an advanced metric especially designed for games and other similar real-time applications.

Spatial and Probabilistic Modelling of Pedestrian Behaviour

This paper investigates the combination of spatial and probabilistic models for reasoning about pedestrian behaviour in visual surveillance systems. Models are learnt by a multi-step unsupervised method and they are used for trajectory labelling and atypical behaviour detection.

An object-based comparative methodology for motion detection based on the F-Measure

The majority of visual surveillance algorithms rely on effective and accurate motion detection. However, most evaluation techniques described in literature do not address the complexity and range of the issues which underpin the design of a good evaluation methodology. In this paper, we explore the problems associated with both the optimising the operating point of any motion detection algorithms and the objective performance comparison of competing algorithms. In particular, we develop an object-based approach based on the F-Measure-a single-valued ROC-like measure which enables a straight-forward mechanism for both optimising and comparing motion detection algorithms. Despite the advantages over pixel-based ROC approaches, a number of important issues associated with parameterising the evaluation algorithm need to be addressed. The approach is illustrated by a comparison of three motion detection algorithms including the well-known Stauffer and Grimson algorithm, based on results obtained on two datasets.

Tracking Human Position and Lower Body Parts Using Kalman and Particle Filters Constrained by Human Biomechanics

In this paper, a novel framework for visual tracking of human body parts is introduced. The approach presented demonstrates the feasibility of recovering human poses with data from a single uncalibrated camera by using a limb-tracking system based on a 2-D articulated model and a double-tracking strategy. Its key contribution is that the 2-D model is only constrained by biomechanical knowledge about human bipedal motion, instead of relying on constraints that are linked to a specific activity or camera view. These characteristics make our approach suitable for real visual surveillance applications. Experiments on a set of indoor and outdoor sequences demonstrate the effectiveness of our method on tracking human lower body parts. Moreover, a detail comparison with current tracking methods is presented.

Finding Paths in Video Sequences

This paper investigates the ta sk of identifying frequently-used pathways from video sequences of natural outdoor scenes. Path models are adaptively learnt from the accumulation of trajectory data over many image frames. Labelled paths are used as an efficient means for compressing the trajectory data for logging purposes. In addition, the path models are used to predict the object’s location many timesteps ahead, and to aid the recognition of unusual behaviour identified as atypical object motion.

Temporal Extension of Laplacian Eigenmaps for Unsupervised Dimensionality Reduction of Time Series

A novel non-linear dimensionality reduction method, called Temporal Laplacian Eigenmaps, is introduced to process efficiently time series data. In this embedded-based approach, temporal information is intrinsic to the objective function, which produces description of low dimensional spaces with time coherence between data points. Since the proposed scheme also includes bidirectional mapping between data and embedded spaces and automatic tuning of key parameters, it offers the same benefits as mapping-based approaches. Experiments on a couple of computer vision applications demonstrate the superiority of the new approach to other dimensionality reduction method in term of accuracy. Moreover, its lower computational cost and generalisation abilities suggest it is scalable to larger datasets.