Ball-and-socket joint pose estimation

Description

This lecture presents a paper on ball-and-socket joint pose estimation using a magnetic field, focusing on an anthropomorphic design with 3-DOF actuated by artificial muscles and tendons. The analysis includes magnetic field design, data-driven joint estimation using LSTM and DVBFs, and experimental results.

Official source

About this result

This page is automatically generated and may contain information that is not correct, complete, up-to-date, or relevant to your search query. The same applies to every other page on this website. Please make sure to verify the information with EPFL's official sources.

Long short-term memory

Long short-term memory (LSTM) network is a recurrent neural network (RNN), aimed to deal with the vanishing gradient problem present in traditional RNNs. Its relative insensitivity to gap length is its advantage over other RNNs, hidden Markov models and other sequence learning methods. It aims to provide a short-term memory for RNN that can last thousands of timesteps, thus "long short-term memory".

Magnetic field

A magnetic field is a vector field that describes the magnetic influence on moving electric charges, electric currents, and magnetic materials. A moving charge in a magnetic field experiences a force perpendicular to its own velocity and to the magnetic field. A permanent magnet's magnetic field pulls on ferromagnetic materials such as iron, and attracts or repels other magnets.

Earth's magnetic field

Earth's magnetic field, also known as the geomagnetic field, is the magnetic field that extends from Earth's interior out into space, where it interacts with the solar wind, a stream of charged particles emanating from the Sun. The magnetic field is generated by electric currents due to the motion of convection currents of a mixture of molten iron and nickel in Earth's outer core: these convection currents are caused by heat escaping from the core, a natural process called a geodynamo.

Highway network

In machine learning, the Highway Network was the first working very deep feedforward neural network with hundreds of layers, much deeper than previous artificial neural networks. It uses skip connections modulated by learned gating mechanisms to regulate information flow, inspired by Long Short-Term Memory (LSTM) recurrent neural networks. The advantage of a Highway Network over the common deep neural networks is that it solves or partially prevents the vanishing gradient problem, thus leading to easier to optimize neural networks.

Magnetic pressure

In physics, magnetic pressure is an energy density associated with a magnetic field. In SI units, the energy density of a magnetic field with strength can be expressed as where is the vacuum permeability. Any magnetic field has an associated magnetic pressure contained by the boundary conditions on the field. It is identical to any other physical pressure except that it is carried by the magnetic field rather than (in the case of a gas) by the kinetic energy of gas molecules.