 I am Reena Rose, Assistant Professor, Department of MCA in Sensavius Engineering College, Nagar Goyal. I know the audience are from different disciplines and I hope that I can make you enjoy my presentation. The topic is role of face recognition in computer vision. By computer vision, we mean making computer identify objects using digital images. So the outline is this, introduction, issues in face recognition, applications, and then how it works and finally the conclusion. Let me first explain what is face recognition system. It's nothing but a computer application that automatically identify or verify person by using facial features from the face. So this particular system has distinct advantage over other biometric systems like fingerprint, palm print and iris extra. So there are a lot of facial features are there like geometric, skin texture, 2D, photometric, etc. I have listed only two. So geometric. To obtain geometric features, we can select some fiducial points from the face and the distances between that point can be the feature. Second feature is skin texture. Textures are nothing but repetitive patterns that can be calculated for every pixel in the, every point in the face with the intensity value difference between it and its surrounding pixels. So you can see an image which makes you understand what is texture. So the art of face recognition is quite complex because of lot of issues like viewing angle of the face. You can see two faces of the same person, first one the frontal face and the second one a rotated one. So from the first, we can have lot of information or the whole information, whole detail of a face. The second, we will get a little fewer than the first one. So how we make computer identify these two faces are for a single person. That's an issue, for lighting. So this can also affect the image. Next is aging. So you can see two images of a single person at the age of two and at the age of 22. So how we make computer identify these two are of same person. Next is facial expressions. So a big smile can render the system less effective. So that is also a serious issue in face recognition. So next is facial faces partially covered with objects like cap, spectacles, etc. So recognizing identical team is also a serious issue that has to be considered while developing face recognition systems. So you can see the features are very similar. So these are all the applications of face recognition. We can prevent frauds in ATM, avoid voters duplication and we can also use face recognition to verify passport visa and driving licenses. We can also use this to identify or verify miscreants at airports, railways stations and malls for better security. We can also use effectively this system effectively for important examinations such as CAD gate, civil service exam, medical and engineering, etc. So we can replace completely the barcode system which is currently present with face recognition. Next we will see how this face recognition system works. So for this first we have to acquire images by using some equipments like CCTV, camera, etc. And before directly processing it we have to do some filtering process because often the equipments used for acquiring images may not produce good images always. So we have to pre-process the images. There are several pre-processing techniques available. We can use that to pre-process the image and after pre-processing we can direct it to feature extraction process. So as I said we can have several features from the face like geometric, skin texture, 3D features and photometric. So after extracting this feature will be the face description of the face recognition system. It is given as the input for face recognition. So this face recognition system has two phases, training phase and testing phase. In the training phase, sample images are collected and they are pre-processed and the features are computed and they are stored in a database. During testing an image under test is pre-processed and the features are extracted and they are matched with the images or the features we have in the database. And the image with the feature that closely matched will be considered as the recognized one. I have listed some recent publications regarding face recognition. Biometric system has distinct advantage over other biometric system because the images are readily, face images are readily available and it can also be easily captured from a distance without the knowledge of the person and there is no interaction needed to verify the person. Comprehensive research is needed to solve outstanding challenges and to propose novel solutions for face recognition system. With this I conclude. Thank you. Rina, you stuck to time and you made a decent presentation.