Project: Improving industrial robotic vision by utilizing embodiment and multi-modal inputs

2019-08-01 – 2023-07-31

Abstract

General purpose
The aim of this project is to develop a cheap and compact robot vision solution capable of 3D positioning
accurate determination of known objects in changing environments and with inexpensive camera hardware, door efficient
use multimodal input and the spatial and temporal embodiment of robots. This system will be existing
surpass robot vision systems by having robots actively monitor camera angles and the flow of incoming data
can be combined with additional domain-specific modalities.