Localization of robot

A robot comprising:

• an input device configured to receive a set of reference images of a surrounding environment of the robot;
• and
• at least one processor configured to:
• associate a set of reference images received by the input device at a specific point in a space with a global position or a global pose of the specific point, and
• store, as training data, the set of reference images and associated global position or global pose of the specific point.

A method of a robot comprising:

• rotating the robot at a specific point for localization;
• receiving a plurality of sequential images based on the rotating of the robot;
• and
• estimating at least one of a position or a pose of the robot in a space, by providing the plurality of sequential images received during the rotation to a trained model based on an artificial neural network.