Juneho Yi

Masked fake face detection using radiance measurements

This research presents a novel 2D feature space where real faces and masked fake faces can be effectively discriminated. We exploit the reflectance disparity based on albedo between real faces and fake materials. The feature vector used consists of radiance measurements of the forehead region under 850 and 685 nm illuminations. Facial skin and mask material show linearly separable distributions in the feature space proposed. By simply applying Fishers linear discriminant, we have achieved 97.78% accuracy in fake face detection. Our method can be easily implemented in commercial face verification systems.

Model-Based 3D Object Recognition Using Bayesian Indexing

This research features the rapid recognition of three-dimensional objects, focusing on efficient indexing. A major concern in practical vision systems is how to retrieve the best matched models without exploring all possible object matches. We have employed a Bayesian framework to achieve efficient indexing of model objects. A decision-theoretic measure of the discriminatory power of a feature for a model object is defined in terms of posterior probability. Domain-specific knowledge compiled off-line from CAD model data is used in order to estimate posterior probabilities that define the discriminatory power of features for model objects. In order to speed up the indexing or selection of correct objects, we generate and verify the object hypotheses for features detected in a scene in the order of the discriminatory power of these features for model objects. Based on the principles described above, we have implemented a working prototype vision system using a feature structure called an LSG (local surface group) for generating object hypotheses. Our object recognition system can employ a wide class of features for generation of object hypotheses. In order to verify an object hypothesis, we estimate the view of the hypothesized model object and render the model object for the computed view. The object hypothesis is then verified by finding additional features in the scene that match those present in the rendered image. Experimental results on synthetic and real range images show the effectiveness of the indexing scheme.

Manufacturing feature recognition toward integration with process planning

Process planning plays a key role by linking CAD and CAM. Its front-end is feature recognition, but feature recognition research has not been in accord with the requirements of process planning. This paper presents an effort for integrating the two activities: feature-based machining sequence generation primarily based on tool capabilities. The system recognizes only manufacturable features by consulting the tool database, and simultaneously constructs dependencies among the features. Then, the A* algorithm is used to search for an optimal machining sequence by the aid of the feature dependencies and a manufacturing cost function.

Structured Light Based Depth Edge Detection for Object Shape Recovery

This research features a novel approach that efficiently detects depth edges in real world scenes. Depth edges play a very important role in many computer vision problems because they represent object contours. We strategically project structured light and exploit distortion of light pattern in the structured light image along depth discontinuities to reliably detect depth edges. Distortion along depth discontinuities may not occur or be large enough to detect depending on the distance from the camera or projector. For practical application of the proposed approach, we have presented methods that guarantee the occurrence of the distortion along depth discontinuities for a continuous range of object location. Experimental results show that the proposed method accurately detect depth edges of human hand and body shapes as well as general objects.

Face Recognition Based on ICA Combined with FLD

Recently in face recognition, as opposed to our expectation, the performance of an ICA (Independent Component Analysis) method combined with LDA (Linear Discriminant Analysis) was reported as lower than an ICA only based method. This research points out that (ICA+LDA) methods have not got a fair comparison for evaluating its recognition performance. In order to incorporate class specific information into ICA, we have employed FLD (Fisher Linear Discriminant) and have proposed our (ICA+FLD) method. In the experimental results, we report that our (ICA+FLD) method has better performance than ICA only based methods as well as other representative methods such as Eigenface and Fisherface methods.

Effective sinogram-inpainting for metal artifacts reduction in X-ray CT images

This research features a new idea for effective sinogram inpainting that boosts the MAR (metal artifacts reduction) performance. We have indentified neighbor pixels that are relevant to the target pixel to be inpainted and only used these pixels in determining an inpainting value. Experimental results show that our methods based on the proposed idea significantly reduce inpainting errors, enhancing the MAR performance. They can be incorporated into any sinogram-inpainting based MAR algorithms.

Using structured light for efficient depth edge detection

This research describes a novel approach that accurately detects depth edges with cluttered inner texture edges effectively ignored. We strategically project structured light and exploit distortion of the light pattern in the structured light image along depth discontinuities to reliably detect depth edges. In practice, distortion along depth discontinuities may not occur or be large enough to detect depending on the distance from the camera or projector. We present methods that guarantee the occurrence of the distortion along depth discontinuities for a continuous range of object location. Experimental results show that the proposed method accurately detects depth edges of shapes of human hands and bodies as well as general objects.

Structural and Functional Brain Connectivity of People with Obesity and Prediction of Body Mass Index Using Connectivity

Obesity is a medical condition affecting billions of people. Various neuroimaging methods including magnetic resonance imaging (MRI) have been used to obtain information about obesity. We adopted a multi-modal approach combining diffusion tensor imaging (DTI) and resting state functional MRI (rs-fMRI) to incorporate complementary information and thus better investigate the brains of non-healthy weight subjects. The objective of this study was to explore multi-modal neuroimaging and use it to predict a practical clinical score, body mass index (BMI). Connectivity analysis was applied to DTI and rs-fMRI. Significant regions and associated imaging features were identified based on group-wise differences between healthy weight and non-healthy weight subjects. Six DTI-driven connections and 10 rs-fMRI-driven connectivities were identified. DTI-driven connections better reflected group-wise differences than did rs-fMRI-driven connectivity. We predicted BMI values using multi-modal imaging features in a partial least-square regression framework (percent error 15.0%). Our study identified brain regions and imaging features that can adequately explain BMI. We identified potentially good imaging biomarker candidates for obesity-related diseases.

A real-time face recognition system using multiple mean faces and dual mode Fisherfaces

This research features an automatic face detection and recognition system. The purpose of the system is for access control to a building or an office. The main feature of the system is face detection and face recognition robust to illumination changes. A novel template matching technique using multiple mean faces (MMF) of various sizes, luminance, and rotations is employed to achieve robust face detection. The system is also capable of operating in two different modes for face recognition: under normal illumination condition and under severe illumination changes.

Face recognition based on locally salient ICA information

ICA (Independent Component Analysis) is contrasted with PCA (Principal Component Analysis) in that ICA basis images are spatially localized, highlighting salient feature regions corresponding to eyes, eye brows, nose and lips. However, ICA basis images do not display perfectly local characteristic in the sense that pixels that do not belong to locally salient feature regions still have some weight values. These pixels in the non-salient regions contribute to the degradation of the recognition performance. We have proposed a novel method based on ICA that only employ locally salient information. The new method effectively implements the idea of ”recognition by parts” for the problem of face recognition. Experimental results using AT&T, Harvard, FERET and AR databases show that the recognition performance of the proposed method outperforms that of PCA and ICA methods especially in the cases of facial images that have partial occlusions and local distortions such as changes in facial expression and at low dimensions.