Computer Vision and Object Motion Engineer

  • Seervision
  • Zürich, ZH, Switzerland
  • 11/02/2020
Full time Data Science Data Engineering Big Data Statistics

Job Description

Seervision aims to be the best in adaptive motion control technology. With this technology, multi-camera robot setups collaborate and can perform all tasks of traditional camera work autonomously, whereby the result of the software-controlled camera can no longer be distinguished from that controlled by a human being. Seervision develops state-of-the-art video analysis algorithms for real-time object recognition. In order to grow, Seervision is in need for a PhD-level interdisciplinary scientist with experience in field of skeletal model predictions and real-time computer vision in broadcast quality camera feeds. Ideally also experienced in neural network training for human detection and movement. With this algorithm, Seervision’s movement estimators will be able to incorporate future skeletal positions to design and optimize future robotic camera trajectories, and using limited data on human movement.

Company: Founded in 2016, Seervision is a multiple award winning spin-off from the Automatic Control Laboratory of ETH Zürich, developed out of research in the field of camera control. All in all, we have carried out more than 20 R&D projects in the field of camera control, as part of more than 5 years of cooperation on various projects and development work for several industrial partners.

Computer Vision and Object Motion Engineer will be: 1) Formulating functional requirements for the skeletal prediction model. 2) Creating a skeletal movement prediction model based on 2D camera frames that must predict the future 3D motion in various cases for at least 2 second in the future. Current objective is to base the predictions on filtered 2D frames captured by the on-set cameras and, in a later stage, achieve semantic action recognition within the same software module. 3) Carrying out comprehensive performance tests of the developed algorithm within hardware. 4) Developing the software module that should be able to predict multiple directions and anticipate on changes in direction. 5) Integrating the prediction software module into the Seervision camera robotics systems.

Education: a PhD-level interdisciplinary scientist in computer science and engineering, mathematics or data science in engineering.

Required skills: To have working experience with GPU programming and deployment of neural networks // Understanding knowledge in the field of adaptive motion control technology // Experience and knowledge of developing algorithms and visualisation // Analysing large-scale datasets and excellent programming skills (Python, MATLAB) // Have experience with the development of neural networks and/or skeletal modelling // Excellent communication skills and enthusiastic, outgoing personality.

Desired skills: Ability to set up evaluation pipelines and data collection for machine learning // Broadly interested and ability to quickly get familiarized with new topics // Ability to work on the same code based with more than 2 direct co-workers an the ability to work in multi-disciplinary teams.

Work environment: Seervision is committed to create a diverse work environment and applicants are considered without regard to ethnicity, religion, gender, sexual orientation or age.

Eligibility criteria: The availability of this position is dependent on the award of the H2020 SME Innovation Associate Grant. The Mobility Rule is applicable to this grant, meaning that at the time of recruitment, the applicant must have not resided or carried out his/her main activity (work. Studies, etc.) for more than a total of 12 months over the last three years in Switzerland.

Are you motivated, innovate an committed to delivering first-rate performance? If you are interested, please contact Damian Frick (jobs@seervision.com).

Additional Information

Eligibility criteria

The availability of this position is dependent on the award of the H2020 SME Innovation Associate Grant. The Mobility Rule is applicable to this grant, meaning that at the time of recruitment, the applicant must have not resided or carried out his/her main activity (work. Studies, etc.) for more than a total of 12 months over the last three years in Switzerland.

Are you motivated, innovate an committed to delivering first-rate performance? If you are interested, please contact Damian Frick (jobs@seervision.com).

Requirements

Offer Requirements

  • REQUIRED EDUCATION LEVELComputer science: PhD or equivalentEngineering: Master Degree or equivalent
  • REQUIRED LANGUAGESENGLISH: Excellent

Skills/Qualifications

Desired skills: Ability to set up evaluation pipelines and data collection for machine learning // Broadly interested and ability to quickly get familiarized with new topics // Ability to work on the same code based with more than 2 direct co-workers an the ability to work in multi-disciplinary teams.

Specific Requirements

Required skills: To have working experience with GPU programming and deployment of neural networks // Understanding knowledge in the field of adaptive motion control technology // Experience and knowledge of developing algorithms and visualisation // Analysing large-scale datasets and excellent programming skills (Python, MATLAB) // Have experience with the development of neural networks and/or skeletal modelling // Excellent communication skills and enthusiastic, outgoing personality.