Motion Fields for Interactive Character Animation

by Danielle Albers on January 23, 2011

in Assignment 1

Assigned paper

  1. The Sentence:
    This paper presents a novel representation of motion data and control that creates motions which appear natural, are highly responsive to real-time user input, and are not explicitly specified in the data. (Synthesized from final sentence of abstract)
  2. The Problem:
    Human motion is a highly-varied and continuous phenomenon: it quickly adapts to different tasks, responds to external disturbances, and in general is capable of continuing locomotion from almost any initial state. This paper focuses on the issue of creating highly agile and realistic interactive locomotion controllers that retain the above motion properties.
  3. Key Idea:
    Uses reinforcement learning and motion fields to provide the ability for a user to control the character in real time and to operate in the fully-continuous configuration space of the character.
  4. What it lets you do:
    a. Key Example:
    Characters can immediately respond to new commands by selecting the optimal motion from the motion field. This promotes agile character motion and requires minimal effort to develop new motion controllers.

    b. Can be used for:
    Create real-time motion controllers, create realistic and responsive virtual characters for gaming and interactive environments.

  5. Resources:
    Video.

Previous post:

Next post: