Image analysis using neural networks for pose and action identification

A method of image analysis to identify human actions represented in an image, comprising:

• analysing an image depicting one or more people using a first computational neural network to determine a set of joint candidates for the one or more people depicted in the image;
• deriving pose estimates from the set of joint candidates that estimate a body configuration for the one or more people depicted in the image;
• and
• identifying a region of interest within the image from the derived pose estimates and analysing the identified region of interest using a second computational neural network to identify an action performed by a person depicted in the image.

An apparatus for performing image analysis to identify human actions from one or more images, comprising:

• a joint-determination module configured to analyse one or more images each depicting one or more people using a first computational neural network to determine for each image a set of joint candidates for the one or more people depicted in the image;
• a pose-estimation module configured to derive for each image pose estimates from the set of joint candidates that estimate a body configuration for the one or more people depicted in the image;
• and
• an action-identification module configured to analyse the derived pose estimates for the one or more images using a second computational neural network and to identify an action performed by a person depicted in the one or more images.