Reference model predictive tracking and rendering

A system comprising:

• memory;
• and
• a processor, coupled to the memory, the memory including instructions, which when executed by the processor, cause the processor to:
• receive at least two images of a user performing a technique;
• determine a position and velocity of a portion of the user from a first to a second image of the at least two images, the second image captured after the first image;
• determine a reference model corresponding to the technique, the reference model including a plurality of frames in a sequence including a first frame corresponding to the first image, the reference model including position and motion data of an avatar;
• searching the plurality of frames to determine, using a nearest neighbor technique, a second frame of the plurality of frames that corresponds to the second image of the at least two images based on the position and velocity of the portion of the user;
• accessing kinematic data from a look-up table related to the second frame;
• and
• output the kinematic data for use in displaying the avatar.

A method comprising:

• receiving a first image of a user performing a technique;
• identifying position and velocity information of the user at a time of capture of the image;
• determining a reference model corresponding to the technique, the reference model including a plurality of frames in a sequence having a first frame including an avatar in a position corresponding to the position of the user in the first image;
• applying, using a processor, the velocity information of the user to the first frame to alter the position of the avatar;
• searching the plurality of frames, using a nearest neighbor technique, to determine a second frame of the plurality of frames corresponding to the altered position of the avatar;
• accessing kinematic data from a look-up table related to the second frame;
• and
• outputting the kinematic data for displaying a second image of the user performing the technique.