The expected value of missing data in a sample taken from a multivariate normal probability distribution is the mean of the conditional distribution of the missing dimensions given the known dimensions. We explain the derivation of this result, demonstrate its application to face image processing, then use it in a new method for recovering shape from image data. The context of our work is the use of 3D facial models to aid in recognition of human faces by humans. We explain the requirement for such models and review the practical possibilities for encoding depth information alongside photographs in identity documents like passports. The best alternative is to derive depths automatically from the photos, as this requires no side information. We show experimentally that conditional density estimation provides accurate face depth recovery, without recourse to explicit modelling of surface shape.

No abstract

No abstract

The MMI-007 project uses both bottom-up and top-down strategies to deal with the difficult problem of object recognition in outdoor scenes. Within this scheme, bottom-up methods fulfil the requirements of providing a concise image description suitable for symbolic reasoning and of providing an initial set of hypotheses to 'bootstrap' the top-down processes. The nature of the imagery does not lend itself easily to the derivation of precise 3-D structural information by optical-flow or stereo techniques. This leads us to call upon a range of static segmentation techniques, each individually capable of capturing some aspect of the diffuse information present in our images. For example, surface, texture and colour homogenity, and boundary smoothness and continuity.

The location of discharges within large transformers and the identification of the nature of the discharge source is a problem. The conventional approach involves the use of a multiplicity of ultrasonic sensors attached to points on the transformer tank chosen to be close to the supposed location of the discharge activity. The output of the ultrasonic sensors is recorded in synchronism with the onset of discharge activity, detected by a sensor coil wound around the transformer's neutral lead. The location of the discharge is estimated according to the differences in time taken for the sound associated with the discharge to reach each transducer.The established approach does not involve analysing the ultrasonic signals, but revolves around the Occurrence of an output from the ultrasonic transducers. The new approach to be described utilises information contained within the signal, to determine the position of the discharge. It relies on the fact that for a given discharge, the ultrasonic signal received by the transducer contains two components, the relative extent of which is determined by the angle between the tank wall and the line joining the transducer to the discharge source. Using this fact, a discharge can be located without having to detect associated electrical activity in the neutral. This makes the location of discharging points much easier, particularly in the electrically noisy environment of a large high voltage substation. Conversely, if the electrical discharge signals are available to synchronise readings from the transducers, then the number of transducer readings required is much reduced for a given accuracy of location. ACOUSTIC EMISSIONAcoustic emission is the process by which sound is produced by rapid energy release within a material. It is possible to listen to this release of energy as the sound wave propagates through the surrounding media. The received waveform depends on the nature of both the source and the materials through which the wave passes. It is a common technique in the field of NonDestructive Testing (NDT) and has found applications in a number of industries [l]. The application of acoustic emission to the location of discharges within a transformer is not new. Commercial systems are available which make use of the acoustic signal in conjunction with electrical signals picked up in the neutral. We have developed a technique which analyses the acoustic signal alone. Tests have been carried out using a laboratory based system with discharges located to a high degree of accuracy.