Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Centre of Research and Technology Hellas INFORMATICS & TELEMATICS INSTITUTE Artificial Intelligence & Information Analysis Group (AIIA) Profile The Informatics and Telematics Institute is a non-profit organization under the auspices of the General Secretarial of Research and Technology of Greece. Since March 10th, 2000 it is a founding member of the Centre of Research and Technology Hellas (CERTH) also supervised by the Greek Secretariat of Research and Technology. The Artificial Intelligence & Information Analysis Group within CERTH/ITI has been active in 2-D/3-D image processing, human-centered interfaces, and multimedia authentication for more than two decades. It is also affiliated with the Aristotle University of Thessaloniki. Profile Two faculty members, several postdoctoral fellows, and more than ten Ph.D. students Participation in 17 European projects (IST, RTN/HCP, ESPRIT, ACTS, LTR/BRA, RACE, TEMPUS, AIM) and 8 national projects Books: 6; One in 3-D Image Processing Algorithms Chapters contributed to edited books: 14 Journal Papers: 111 Conference Papers: 273 Role Prior experience in topics related to the project: 3-D Image Processing & Graphics Human-centered interfaces Face analysis; facial feature extraction; face tracking; facial feature tracking; face expression recognition; Graphical communication: audio-visual speech analysis; Role Research activities in synthesis tasks Development of talking heads (virtual salesmen) with emphasis in Texture mapping; Synthesis of facial expressions; prototypes; Synthesis compliant with standards MPEG-4 FDPs, FACS, etc. Development activities: Contribution to the integration of generic tools for graphic design, and speech recognition /synthesis into the Worlds Studio Platform. State of the art in Facial Modeling and Animation 2D and 3D morphing Manually corresponding features Combination of 2D morphing with 3D geometric transformations Physics based muscle modeling Spring Mesh Muscle Vector Muscle Layered Spring Mesh Muscle Finite Element Method State of the art in Facial Modeling and Animation Pseudo or Simulated Muscle Free Form Deformations Combination of 2D morphing with 3D geometric transformations Wires Model Fitting Adaptation to an example face MPEG-4 Compatible Models Facial Action Coding System Example-driven deformations of the model Speech driven heads Face Detection using Color Information HSV thresholding based on facial colors Connected components labeling and analysis Segmentation of the image Moments computation Best fit ellipse image Spatial constraint for face contour Initialization of the snake Snake deformation by energy minimization Inner face contour image HSV Thresholding & Connected Components Labeling Thresholding based on facial colors (segmentation) Keep only the pixels having color similar to facial texture Initial image Segmented image HSV thresholding Connected components labeling Moments Computation Ellipses in the segmented image Best fit ellipse Segmented image Best Fit Ellipse image Moments Computation Model Superposition on Face Images ICP for superposition of the model on points already defined on a face image Mass-Spring Models to fit the face model on the face image I(terative) C(losest) P(oint) algorithm ICP is based on the Closest Set of Points Closest Set of Points leads to the Quaternion Quaternion an easily handled vector the basis of the ICP transformations similar to the rotation and translation matrices Convergence Monotonically to the nearest local minimum rapid during the first few iterations globallity depends on the initial parameters Mass-Spring Models FEM restricted models Simulates models as masses connected with springs Physics based simulation four masses connected among themselves with uniform springs Examples Randomly initial positioning of the face model Interactive definition of points on the face image Examples Fit of the model by applying the ICP algorithm Fit of the model by applying the Mass-Spring Model Ellipse fitting Ellipse determination based on the model’s position Model Superposition Ellipse Image Model Superposition based on model’s contour Combination of two methods Ellipse extracted using Moments Computation Ellipse extracted using the Model Fitting procedure Intermediate ellipse (scale) Intermediate Ellipse image Intermediate ellipse Spatial Constraint for Face Contour Initialization of the snake Circular deformable model Intermediate Ellipse Initialized Snake Spatial Constraint for Face Contour Snake Deformation & Face Contour Detection Snake Deformation by Energy Minimization Snake Energy: Internal Energy: External Energy: Etotal Eint Eext Eint (i ) i i 1 i i 1 i 1 2i i 1 2 Eext (i ) I ( x, y) k 1 2 2 d (i 1 ,i ) d (i ,i 1 ) Intermediate Ellipse Deformated Snake Snake Deformation Future Work Statistical analysis/synthesis of the images & facial models Eigen-decomposition and PCA of the features Features correspondences of the images & the model Extraction of Face Definition Points (MPEG-4) Extraction of Facial Expressions