DECODE

Continual learning for visual and multi-modal encoding of human surrounding and behavior

Continual learning for visual and multi-modal encoding of human surrounding and behavior

Machine Learning, and in particular Artificial Intelligence (AI) in Deep Learning, has revolutionized Computer Vision in almost all areas. These include topics such as motion estimation, object recognition, semantic segmentation (division and classification of parts of an image), pose estimation of people and hands, and many more. A major problem with this method is the distribution of the data. Training data often differs greatly from real applications and do not adequately cover them. Even if suitable data are available, extensive retraining is time-consuming and costly. Adaptive methods that continuously learn (lifelong learning) are the central challenge for the development of robust, realistic AI applications. In addition to the rich history in the field of general continuous learning, the topic of continuous learning for machine vision under real conditions has recently gained interest. The goal of the DECODE project is to explore continuously adaptive models for reconstructing and understanding human motion and the environment in application-related environments. For this purpose, mobile, visual and inertial sensors (accelerometers and angular rate sensors) will be used. For these different types of sensors and data, different approaches from the field of continuous learning will be researched and developed to ensure a smooth transfer from laboratory conditions to everyday, realistic scenarios. The work will concentrate on in the areas of segmented image and video segmentation, kinematic and pose estimation and the estimation of kinematics and pose of the human body as well as the representation of movements and their context. The field of potential applications for the methods developed in DECODE is wide-ranging and includes detailed ergonomic analysis of human-machine analysis of human-machine interactions, for example in the workplace, in factories, or in vehicles.

Contact

Dr.-Ing. Nadia Robertini

Dr.-Ing. René Schuster

GreifbAR

Greifbare Realität - geschickte Interaktion von Benutzerhänden und -fingern mit realen Werkzeugen in Mixed-Reality Welten

Greifbare Realität – geschickte Interaktion von Benutzerhänden und -fingern mit realen Werkzeugen in Mixed-Reality Welten

On 01.10.2021, the research project GreifbAR started under the leadership of the DFKI (research area Augmented Reality). The goal of the GreifbAR project is to make mixed reality (MR) worlds, including virtual (VR) and augmented reality (“AR”), tangible and graspable by allowing users to interact with real and virtual objects with their bare hands. Hand accuracy and dexterity is paramount for performing precise tasks in many fields, but the capture of hand-object interaction in current MR systems is woefully inadequate. Current systems rely on hand-held controllers or capture devices that are limited to hand gestures without contact with real objects. GreifbAR solves this limitation by introducing a sensing system that detects both the full hand grip including hand surface and object pose when users interact with real objects or tools. This sensing system will be integrated into a mixed reality training simulator that will be demonstrated in two relevant use cases: industrial assembly and surgical skills training. The usability and applicability as well as the added value for training situations will be thoroughly analysed through user studies.

Partners

Berliner Charite (University Medicine Berlin) NMY (Mixed reality applications for industrial and communication customers) Uni Passau (Chair of Psychology with a focus on human-machine interaction).

Contact

Dr. Dipl.-Inf. Gerd Reis

Dr.-Ing. Nadia Robertini