Description du projet
Des caméras autonomes gardent les yeux sur leur cible
Les progrès étonnants de la qualité des appareils photo des smartphones au cours de la dernière décennie témoignent de la révolution du matériel photo et vidéo, stimulée en grande partie par les réseaux sociaux et les jeux. Les progrès du matériel ne représentent bien sûr qu’une des facettes de l’équation permettant à chacun de prendre de meilleures photos et vidéos. Le projet SeerPredict, financé par l’UE, développe une solution logicielle pour aider les professionnels à faire encore un meilleur travail. Un logiciel qui apprend à prédire les mouvements du squelette en offrant jusqu’à deux secondes d’anticipation, en se basant sur l’apprentissage automatique de séquences squelettiques, permettra aux systèmes de caméra autonomes de maintenir la mise au point sur leurs cibles.
Objectif
Seervision was founded in 2016, as a spin-off from the ETH Zürich and developed out of the Automatic Control Laboratory research group in the field of camera control. With their broad experience in the field of computer vision, machine learning and artificial intelligence, Seervision aims to develop fully autonomous camera systems based on vision-based 3D motion capturing. The goal is to predict skeletal motion for maximum of 2 seconds in future out of skeletal kinematics and 2D camera frames. Therefore a prediction model has to be developed with a machine learning algorithm to train the model with data on skeletal motion. Using a trained prediction model for skeletal motion based on 3D vision-based motion capturing, Seervision aims to control their camera systems fully autonomously. Currently, there are no such models commercially available and the level of autonomy in commercially available camera systems is much lower.
The proposed recruitment of an Innovation Associate (IA) in this project will have significant impact on the business opportunities of Seervision by bringing in the lacking expertise on skeletal motion and development of prediction models based on vision-based motion capturing. The SME IA grant will help to overcome the recruitment barriers of being a relatively unknown, small enterprise that cannot compete with the larger companies due to limited visibility, salary and recruitment budget.
Seervision wants to recruit a Computer Vision and Object Motion Engineer that understands the principles of 3D skeletal motion and is able to model and deploy machine learning algorithms within this context. The main added value of this opportunity for the IA is to pursue the transition from an academic environment focusing on basic research to a business environment focusing on developing competitive and cutting-edge products. Next to this, the IA will get the chance to influence and participate in the early stages of a young and innovative company.
Champ scientifique
- engineering and technologyelectrical engineering, electronic engineering, information engineeringelectronic engineeringsensorsoptical sensors
- natural sciencescomputer and information sciencesartificial intelligencecomputer vision
- natural sciencescomputer and information sciencesartificial intelligencemachine learning
- natural sciencescomputer and information sciencessoftwaresoftware applicationssimulation software
- natural sciencesmathematicsapplied mathematicsmathematical model
Programme(s)
Régime de financement
CSA-LSP - Coordination and support action Lump sumCoordinateur
8092 ZURICH ETH-ZENTRUM
Suisse
L’entreprise s’est définie comme une PME (petite et moyenne entreprise) au moment de la signature de la convention de subvention.