Are you excited about creating a digital twin of the 3D world around you? The next generation of autonomous cars, robots or mixed reality devices will require an efficient and scalable learned 3D representation of the scene which can jointly serve a multitude of downstream tasks.

Desirable goals and possible research directions are:

• learned online scene updates from input data streams, collaborative/crowd-based updating (sequence to vector learning) of a scene representation;
• rich multi-modal scene representations (geometry, texture, material, semantics, object relations, usage information, etc.);
• self-supervised/weakly supervised learning with input data reproduction via differentiable rendering techniques;
• efficient data structures for learned scene encodings;
• multi-sensor data fusion.

Autonomous agents need to process large amounts of data and keep previously seen information in memory in an efficient and compressed manner. For example projects which perform some first steps check out the references down below [1-6].

This fully-funded PhD position is within the Computer Vision Lab at the informatics institute of UvA and ATLAS lab - a collaboration between UvA and TomTom. The goal of ATLAS lab is the development new machine learning-based algorithms for high definition map creation for self-driving vehicles. The ATLAS lab is also part of the Innovation Center for Artificial Intelligence (ICAI) such that there are plenty of possibilities to network and benefit from the participation of related events.

What are you going to do?

• Design and implement novel algorithms/data structures for learned 3D scene representations.
• You will work at the intersection of 3D computer vision, (inverse) graphics, machine learning, and optimization to develop next generation mapping and scene understanding algorithms.
• Collaborate with other researchers within the lab and TomTom.
• Pursue and complete a PhD thesis within the appointed duration of four years.
• Present research results at international conferences and journals.
• Assist in teaching activities such as lab assistance and student supervision.

What do we require of you?

• A Master’s degree in Artificial Intelligence, Computer Science, Mathematics, or a closely related field.
• A strong scientific and mathematical background in deep learning and computer vision.
• An interest, passion, self-motivation and excitement in developing new algorithms for large-scale scene understanding.
• Good programming skills, experience with C++/Python and deep learning frameworks (PyTorch/Tesnorflow/JAX).
• A good academic record and eagerness to tackle core scientific problems.
• Fluent in English, both written and spoken.

Our offer

A temporary contract for 38 hours per week for the duration of four years (the initial contract will be for a period of 18 months and after satisfactory evaluation it will be extended for a total duration of four years). This should lead to a dissertation (PhD thesis). We will draft an educational plan that includes attendance of courses and (international) meetings. We also expect you to assist in teaching undergraduates and Master students.

The salary, depending on relevant experience before the beginning of the employment contract, will be €2,443 to €3,122 (scale P) gross per month, based on a fulltime contract (38 hours a week). This is exclusive 8% holiday allowance and 8.3% end-of-year bonus. A favourable tax agreement, the ‘30% ruling’, may apply to non-Dutch applicants. The Collective Labour Agreement of Dutch Universities (CAO NU) is applicable.

Are you curious about our extensive package of secondary employment benefits like our excellent opportunities for study and development? Take a look here.

References

[1] NeuralFusion: Online Depth Fusion in Latent Space,
Silvan Weder, Johannes L. Schönberger, Marc Pollefeys, and Martin R. Oswald.
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2021

[2] DeepSurfels: Learning Online Appearance Fusion,
Marko Mihajlovic, Silvan Weder, Marc Pollefeys, and Martin R. Oswald.
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2021

[3] NVS-MonoDepth: Improving Monocular Depth Prediction with Novel View Synthesis,
Zuria Bauer, Zuoyue Li, Sergio Orts-Escolano, Miguel Cazorla, Marc Pollefeys, and Martin R. Oswald. International Conference on 3D Vision (3DV), 2021

[4] Shape from Blur: Recovering Textured 3D Shape and Motion of Fast Moving Objects,
Denys Rozumnyi, Martin R. Oswald, Vittorio Ferrari, and Marc Pollefeys. NeurIPS, 2021

[5] RoutedFusion: Learned Real-time Depth Map Fusion,
Silvan Weder, Johannes L. Schönberger, Marc Pollefeys, and Martin R. Oswald.
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2020

[6] KAPLAN: A 3D Point Descriptor for Shape Completion,
Audrey Richard, Ian Cherabier, Martin R. Oswald, Marc Pollefeys, and Konrad Schindler.
International Conference on 3D Vision (3DV), 2020

Faculty of Science - Informatics Institute

The Faculty of Science has a student body of around 7,000, as well as 1,600 members of staff working in education, research or support services. Researchers and students at the Faculty of Science are fascinated by every aspect of how the world works, be it elementary particles, the birth of the universe or the functioning of the brain.

The mission of the Informatics Institute is to perform curiosity-driven and use-inspired fundamental research in Computer Science. The main research themes are Artificial Intelligence, Computational Science and Systems and Network Engineering. Our research involves complex information systems at large, with a focus on collaborative, data driven, computational and intelligent systems, all with a strong interactive component.