Motor learning lies at the heart of how humans and animals acquire their skills. Understanding of this process enables many benefits in Robotics, physics-based Computer Animation, and other areas of science and engineering. In this thesis, we develop a computational framework for learning of agile, integrated motor skills. Our algorithm draws inspiration from the process by which humans and animals acquire their skills in nature. Specifically, all skills are learned through a process of staged, incremental learning, during which progressively more complex skills are acquired and subsequently integrated with prior abilities. Accordingly, our learning algorithm is comprised of three phases. In the first phase, a few seed motions that accomplish goals of a skill are acquired. In the second phase, additional motions are collected through active exploration. Finally, the third phase generalizes from observations made in the second phase to yield a dynamics model that is relevant to the goals of a skill. We apply our learning algorithm to a simple, planar character in a physical simulation and learn a variety of integrated skills such as hopping, ipping, rolling, stopping, getting up and continuous acrobatic maneuvers. Aspects of each skill, such as length, height and speed of the motion can be interactively controlled through a user interface. Furthermore, we show that the algorithm can be used without modification to learn all skills for a whole family of parameterized characters of similar structure. Finally, we demonstrate that our approach also scales to a more complex quadruped character.

(далее…)

Категория: Научные статьи | Нет комментариев »

A method and system are provided for the visual display of anatomical forces, that system having: a motion capture system; a computer, receiving data from said motion capture system; and a computational pipeline disposed on said computer; that computational pipeline being configured to calculate muscle forces and joint torques in real time and visually display those forces and torques.

(далее…)

Категория: Патенты | Нет комментариев »

The sensing organs are exponentially better than any of analogous artificial ones. That is why using them in full scale is a perspective trend to the efficient (advanced) machine perception. On the other hand, limitations of sensing organs could be replaced by the perfect artificial ones with the subsequent training the nervous system on their output signals. An attempt to lay down the foundations of biosensing by natural sensors and in addition to them by the artificial trans-ducers of physical quantities, also with their expansion into space arrays and external/implantable functioning in rela-tion to the nervous system is performed. The advances in nanotechnology are opening the way to achieving direct elec-trical contact of nanoelectronic structures with electrically and electrochemically active neurocellular structures. The transmission of the sensors’ signals to a processing unit has been maintaining by an electromagnetic transis-tor/memristor (externally) and super-conducting transducer of ionic currents (implantable). The arrays of the advanced sensors give us information about the space and direction dynamics of the signals’ spreading.The measuring method and necessary performance data of the sensor for the robot’s orientation in the ambient magnetic field with living be-ing-machine interaction in order to obtain input and output signals from brain and motor nerves to the measurement system and vice versa are introduced. The range of applied sensors differs from an induction sensor to superconducting induction magnetometer. The analytical expressions for arrangements of the head sensors in differential and vector (3D) relative positions are deduced. Sensitivity of the perception method makes it possible to recognize the linear translation of 10−2 m and disposal in space of 10−3 m3. Interaction between living beings and robotic equipment is given analytical treatment.

(далее…)

Категория: Научные статьи | Нет комментариев »

Numerous medical and biomechanical applications invol-ve electromyogram (EMG) signal processing in 19 real time. Amplitude analysis of the EMG often requires computation of the signal’s linear envelope. 20 For this purpose, several methods are commonly described in the literature; however, not all match 21 the speed requirement of real-time applications. We introduce an implementation which accelerates 22 the computation of EMG signals linear envelopes, based on the pipe-line commonly found in the literature 23 for this kind of operation. The algorithm improves the computa-tion’s time requirement, at the expense of 24 memory requirement, by using the result of the envelope’s computation at the previous instant. This 25 algorithm saves approximately 96% of the computation time and allows computing linear envelopes of 26 several EMG signals in real time.

(далее…)

Категория: Научные статьи | Нет комментариев »

In this study, we focus on a rehabilitation motion of
human wrist joint and aim at supporting these rehabilitation training by introducing a mechanical system. A pneumatic parallel manipulator is introduced since it can drive enough D.O.F to correspond to a wrist motion and has inherent compliance characteristics resulted from air compressibility. We propose a new training method for a muscle strengthening training based on ElectroMyoGraphy (EMG), where a payload is set between joint angle and the muscle we want to train directly. In addition, we propose a method to de-tect muscle fatigue based on a frequency analysis of EMG. The effectiveness of the proposed method is confirmed through some experiments.

(далее…)

Категория: Научные статьи | Нет комментариев »