T. J. Stonham. Practical face recognition and verification with WISARD. In Hadyn Ellis, Malcolm Jeeves, Freda Newcome, and Andy Young, editors, Aspects of Face Processing, pages 426--441. Martinus Nijhoff, Dordrecht, 1986.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....However, there is nothing in the basic design which requires this, and in section 8 we show how FindFace can be adapted to perform in a different context. Also in this section we indicate some of the applications of this work. 3 2 Background An early successful face recogniser is WISARD [Sto86]. During the initial learning phase, many different instances of the face of an individual are presented, and recognition subsequently occurs when a new image is sufficiently similar to one of the training instances. Subsequent approaches have been based on neural net techniques [CF90] OA88] ....

T. J. Stonham. Practical face recognition and verification with WISARD. In Hadyn Ellis, Malcolm Jeeves, Freda Newcome, and Andy Young, editors, Aspects of Face Processing, pages 426--441. Martinus Nijhoff, Dordrecht, 1986.

....characteristic of flexibility, simplicity and efficiency of execution. This paper will first give a brief overview of how salmon biomass could be estimated using underwater stereo imaging techniques. Secondly a short review of the n tuple classifier, which is based on the WISARD architecture [8] and how it can be fitted into the frame work of the salmon biomass estimation project will be described. Finally experiments are described on how the performance of the classifier is tested according to the requirements of the application and how well the n tuple classifier can be used as a tool ....

....techniques can be employed BMVC99 341 for this task, however these methods will often required large amount of computation resources. The authors have suggested the problems could be overcome by the use of an n tuple classifier, which is based on the implementation of the WISARD architecture [8]. Provisional experimental results are promising and further experimental results have shown that the classifier could be used to solve the problems in question. 3 Image Segmentation The grey level image in figure 2 is too complicated for the classifier to be applied to. Therefore the images are ....

[Article contains additional citation context not shown here]

Stonham, IJ. Practical face recognition and verification with WISARD. Aspects of face processing, Nartinus Nighoff, 1986. BMVC99 349

....work is either recognition by using facial profiles (for example, see [23] 24] or using the frontal views. In this paper we are interested in the latter case where the input is an intensity image of the frontal view of a face. Previous related work can be found in [25] the WISARD system ( 26] [27]) and the dynamic link architecture for face recognition [28] One of the early systems built for this task is described in [29] The system automatically localizes features such as corners of the eyes, nostrils, mouth etc. Then a set of sixteen facial parameters corresponding to these features is ....

J. Stonham, "Practical face recognition and verification with wisard," in Aspects of Face Processing (F. N. H. Ellis, M. Jeeves and A. Young, eds.), Dordrecht: Martinus Nijhoff, 1986.

....different approaches to compare objects and in particular faces. The first one, called global or holistic considers the face image as a whole without trying to analyse its content. The utilized techniques encompass pixel correlation, artificial neural network with images as input pattern (see [Kohonen89,Stonham86]) or image transform such as the principal component analysis [Turk90] to name a few. The same techniques are advantageously applied to facial parts, usually rectangular areas around the nose, the eyes and the mouth ( Brunelli93,Beymer94] The second approach, called featural , looks for ....

T. Stonham, PRACTICAL FACE RECOGNITION AND VERIFICATION WITH WISARD, In Aspects of Face Processing, Doordrecht, Netherlands: Martinus Nijhoff Publishers, pp 426-441, 1986.

....recognition. The recent results from the various neural network approaches that have been applied toward automated facial expression recognition show that a lot of progress has been made since the conception of the early neural network, a perceptron, that could classify smiles from frowns (Stonham 1986). These recent attempts suggest that facial expression recognition using a computational approach appears quite feasible. Given that facial expressions have been associated with inner emotional states, we now explain how an automatic facial expression interpreter could be integrated in a larger ....

Stonham, T.J. 1986. "Practical face recognition and verification with Wisard". In H. Ellis and M.A. Jeeves (eds), Aspects of Face Processing. Lancaster: Martinus Nijhoff, 426-441.

.... has attracted much publicity is WISARD (Wilkie, Stonham and Aleksander s Recognition Device) a general object recognition system which is based on neural network principles, implemented by massive dedication of memory to the storage of responses to instances of patterns on which it is trained [40, 41]. WISARD is a pattern recognition machine with a special semi parallel structure, using statistical pattern classification, and self adapting, which can be used for face recognition and verification. In a face recognition test the system was set up for 16 individuals, with the input resolution of ....

T.J. Stonham. Practical face recognition and verification with WISARD. In Aspects of Face Processing, pages 426 -- 441. Martinus Nighoff, 1986.

.... et al. 11] and Hong[19] and vector quantization (Ramsay, et al. 31] Connectionist approaches to face recognition also use pictorial representations for faces (Kohonen[24] Fleming and Cottrell [16] Edelman, Reisfeld, and Yeshurun[15] Weng, Ahuja, and Huang[40] Fuchs and Haken[17] Stonham[36]) Since the networks used in connectionist approaches are just classifiers, these approaches are similar to the ones described above. Different pixel based representations have been used, with [24] 16] 17] using the original grey level images. 40] uses directional edge maps, 36] uses a ....

....Stonham[36] Since the networks used in connectionist approaches are just classifiers, these approaches are similar to the ones described above. Different pixel based representations have been used, with [24] 16] 17] using the original grey level images. 40] uses directional edge maps, [36] uses a thresholded binary image, and [15] uses Gaussian units applied to the grey level image. Hybrid representations that combine the geometrical and pictorial approaches have been explored, such as Cannon et al. 9] whose feature vector face representation includes geometrical and ....

T.J. Stonham. Practical face recognition and verification with WISARD. In M. Jeeves, F. Newcombe, and A. Young, editors, Aspects of Face Processing, pages 426--441. Martinus Nijhoff Publishers, Dordrecht, 1986.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC