PhD Scholarships in Computer Vision with Deep Learning for Human-Machine Collaboration -Technical University of Denmark

Related Articles

UNIL Master’s grants

Eric Bleumink Scholarship for Developing Countries at the University of Groningen – Netherlands

Bourse de recherche Eduard Gübelin dans les domaines des sciences de la Terre, Suisse

Do you want to advance the state-of-the-art research in computer vision systems using human-machine collaboration? The Technical University of Denmark (DTU) and the Pioneer Centre for Artificial Intelligence invites applicants for two fully-funded 3-year PhD positions. The positions are funded by the Sapere Aude: DFF-Starting grant. The positions are offered at DTU, in the Department of Computer Science and Applied Mathematics, at the Visual Computing Group, and will work closely with researchers from the Pioneer Centre for AI.

Project description

Learning visual recognition models requires huge amounts of data, especially when relying on data-hungry deep neural networks. These data need to be manually and passively annotated by humans. This labeling procedure is one of the most important steps in a machine learning pipeline and is tedious, expensive, and susceptible to noise leading to erroneous labels. Our goal is to establish a deeper human-machine collaboration for all stages of the learning pipeline via two PhD projects:

• Project A: Human-machine labeling. In this project, you will first develop an active labeling and learning scheme where each step of the learning procedure is actively assisted by a learning algorithm. Then, you will advance the capabilities of current labeling schemes towards a deeper human-machine collaboration. This PhD project will be co-supervised by Associate Professor Dimitrios Papadopoulos (DTU) and Professor Serge Belongie (KU).
• Project B: Multimodal labeling and learning. The goal of this project is to leverage information from different modalities that naturally occur near web images to assist the image annotation.  You will develop state-of-the-art multimodal models for self-supervision with human-in-the-loop text supervision. This PhD project will be supervised by Associate Professor Dimitrios Papadopoulos (DTU). 

Responsibilities

During the PhD program, you are expected to:

• Work with state-of-the-art deep learning computer vision models.
• Develop state-of-the-art models for human-machine collaboration in computer vision.
• Publish scientific papers and present research results in top computer vision conferences (CVPR, ECCV, ICCV).
• Spend a few months during the PhD program abroad.
• Work as a teaching assistant in courses at DTU. 

Qualifications

Candidates should have the following required skills:

• A master’s degree (120 ECTS points) or a similar in Computer Science, Informatics, Computer Engineering, Applied Mathematics or equivalent.
• Proven experience in Python.
• Hands-on experience with deep learning frameworks.
• High level of motivation and innovation.
• Experience in computer vision is a plus.
• Excellent written and communication skills in English. 

Salary and appointment terms

The appointment will be based on the collective agreement with the Danish Confederation of Professional Associations. The allowance will be agreed upon with the relevant union. The positions are full-time positions. The period of employment is 3 years; start in summer 2023 or as soon as possible thereafter.

Further information

If you have questions, you are very welcome to contact Associate Professor Dimitrios Papadopoulos at dimp@dtu.dk.

If you are applying from abroad, you may find useful information on working in Denmark and at DTU at DTU – Moving to Denmark. Furthermore, you have the option of joining our monthly free seminar “PhD relocation to Denmark and startup “Zoom” seminar” for all questions regarding the practical matters.

Application procedure

To apply, please read the full job PhD Scholarships in Computer Vision with Deep Learning for Human-Machine Collaboration

Deadline: 15 May 2023 (Danish time).