In the production of the CG movie for Resident Evil, Capcom's renowned game title reborn for Nintendo's GameCube, we used Eyematic Interfaces' FaceTracker combined with Vicon8 optical motion capture system. Using the optical motion capture system and the markers on the actor's face, technically we can capture the facial motion data to create facial expressions. We can create the smooth and real facial expressions that we could not get through using the key frame animation technique. In the facial expressions, however, the eyes are the most important but it was obviously difficult to put the markers on the actor's eyeballs. We solved this problem and completed the facial capture of Resident Evil using the Eyematic Interfaces' FaceTracker combined with VMS's Vicon8 motion capture system. With Eyematic's FaceTracker, based on the image recognition technique from video, we could obtain the motion capture data in real time without any markers on the face. However, FaceTracker can process only the facial images shot from the front. In the motion capture session, we wanted the actors to have freedom to move as his actions would dictate. In order to solve this problem, we designed the head mount video camera to use with FaceTracker. We combined a small video camera and white LEDs for facial lighting; this had to be as light and as small as possible so that the head mount wouldn't create an obstacle for the Vicon8 facial capturing the detailed expressions. We were concerned about a possible problem with interference between the FaceTracker and Vicon8; fortunately it was not a problem. Head mount camera for FaceTracker and facial markers for Vicon8. FaceTracker extracted the control points (including iris of the eyeballs) from the video images. The intermediate model with control points. Usually the captured facial data will be adapted to the primary control points on the CG character's face. This process is time consuming, in general, and needs to be reprocessed whenever the CG model design is modified. To solve the problem, we created a simple intermediate model between the motion capture data and final CG model. We can control the CG model through the intermediate model without the usual difficulty caused by design changes. We could not capture all the detail of the facial expressions like lines on the forehead and eyelids. We created lines calculated from the markers' position based on muscle movements, and we created eyelid movement by morphing key expressions in which we control the morphing parameter using the markers' position data. The final facial image with the additional expressions. Capturing the torso and facial motion together. In the CG animation, it is important to capture the natural expressions while capturing the actor on the stage. In the motion capture session of Resident Evil, we could allow the actor to turn his body and face up to 90 degrees in the combination with body capture. We are now able to capture facial expressions, and hence, data while turning freely through 360 degrees. MoCap Actor: Tetsuya Matsui (Matsui Enterprises Inc.) PDF Download : bio_maiking_E.pdf click to download Acrobat Reader for this PDF files.