A. Jonathan Howell

Face recognition using radial basis function neural networks

This paper presents experiments using an adaptive learning compo nent based on Radial Basis Function RBF networks to tackle the unconstrained face recognition problem using low resolution video in formation Firstly we performed preprocessing of face images to mimic the e ects of receptive eld functions found at various stages of the hu man vision system These were then used as input representations to RBF networks that learnt to classify and generalise over di erent views for a standard face recognition task Two main types of preprocessing Di erence of Gaussian ltering and Gabor wavelet analysis are com pared Secondly we provide an alternative face unit RBF network model that is suitable for large scale implementations by decomposi tion of the network which avoids the unmanagability of neural net works above a certain size Finally we show the D shift scale and y axis rotation invariance properties of the standard RBF network Quantitative and qualitative di erences in these schemes are described and conclusions drawn about the best approach for real applications to address the face recognition problem using low resolution images

Learning identity with radial basis function networks

Radial basis function (RBF) networks are compared with other neural network techniques on a face recognition task for applications involving identification of individuals using low-resolution video information. The RBF networks are shown to exhibit useful shift, scale and pose (y-axis head rotation) invariance after training when the input representation is made to mimic the receptive field functions found in early stages of the human vision system. In particular, representations based on difference of Gaussian (DoG) filtering and Gabor wavelet analysis are compared. Extensions of the techniques to the case of image sequence analysis are described and a time delay (TD) RBF network is used for recognising simple movement-based gestures. Finally, we discuss how these techniques can be used in real-life applications that require recognition of faces and gestures using low-resolution video images.

Invariance in radial basis function neural networks in human face classification

This paper is concerned with the types of invariance exhibited by Radial Basis Function (RBF) neural networks when used for human face classification, and the generalisation abilities arising from this behaviour. Experiments using face images in ranges from face-on to profile show the RBF networks invariance to 2-D shift, scale and y-axis rotation. Finally, the suitability of RBF techniques for future, more automated face classification purposes is discussed.

RBF Network Methods for Face Detection and Attentional Frames

In this paper we introduce a set of adaptive vision techniques which could be used, for example, in video-conferencing applications. First, we present methods for finding faces and selecting attentional frames to focus visual processing. Second, we present methods for recognising individual gesture phases for camera control. Finally, we discuss how these techniques can be extended to ‘virtual groups’ of multiple people interacting at multiple sites.

Learning Gestures for Visually Mediated Interaction

This paper reports initial research on supporting Visually Mediated Interaction (VMI) by developing person-specific and generic gesture models for the control of active cameras. We describe a time-delay variant of the Radial Basis Function (TDRBF) network and evaluate its performance on recognising simple pointing and waving hand gestures in image sequences. Experimental results are presented that show that high levels of performance can be obtained for this type of gesture recognition using these techniques, both for particular individuals and across a set of individuals. Characteristic visual evidence can be automatically selected and used even to recognise individuals from their gestures, depending on the task demands.

Recognising Simple Behaviours Using Time-Delay RBF Networks

This paper presents experiments using a radial basis function variant of the time-delay neural network with image sequences of human faces. The network is shown to be able to learn simple behaviours based on y-axis head rotation and generalise on different data. The network models suitability for future dynamic vision applications is discussed.

Integrated vision system for the semantic interpretation of activities where a person handles objects

Interpretation of human activity is primarily known from surveillance and video analysis tasks and concerned with the persons alone. In this paper we present an integrated system that gives a natural language interpretation of activities where a person handles objects. The system integrates low-level image components such as hand and object tracking, detection and recognition, with high-level processes such as spatio-temporal object relationship generation, posture and gesture recognition, and activity reasoning. A task-oriented approach focuses processing to achieve near real-time and to react depending on the situation context.

Gesture Recognition for Visually Mediated Interaction

This paper reports initial research on supporting Visually Mediated Interaction (VMI) by developing person-specific and generic gesture models for the control of active cameras. We describe a time-delay variant of the Radial Basis Function (TDRBF) network and evaluate its performance on recognising simple pointing and waving hand gestures in image sequences. Experimental results are presented that show that high levels of performance can be obtained for this type of gesture recognition using such techniques, both for particular individuals and across a set of individuals. Characteristic visual evidence can be automatically selected, depending on the task demands.

Learning Gestures with Time-Delay RBF Networks

This paper present experiments using an radial basis function variant of the time-delay neural network with image sequences of simple gestures. The network is shown to be able to learn gestures such as pointing left or right and waving and generalise on different data.

Developing Context Sensitive HMM Gesture Recognition

We are interested in methods for building cognitive vision systems to understand activities of expert operators for our ActIPret System. Our approach to the gesture recognition required here is to learn the generic models and develop methods for contextual bias of the visual interpretation in the online system. The paper first introduces issues in the development of such flexible and robust gesture learning and recognition, with a brief discussion of related research. Second, the computational model for the Hidden Markov Model (HMM) is described and results with varying amounts of noise in the training and testing phases are given. Third, extensions of this work to allow both top-down bias in the contextual processing and bottom-up augmentation by moment to moment observation of the hand trajectory are described.