[visionlist] PhD Scholarship Position in 3D Representation Learning

[emap3088 at gmail.com]

At the department of Electrical Engineering research and education are performed in the areas of Communication and Antenna systems, Systems and Control, Computer vision, Signal processing and Biomedical engineering, and Electric Power Engineering. Our knowledge is of use everywhere where there is advanced technology with integrated electronics. We work with challenges for a sustainable future in society of today, for example in the growing demands concerning efficient systems for communications and electrifying.

We offer a dynamic and international work environment with about 200 employees from more than 20 countries, and with extensive national and international research collaborations with academia, industry and society. 



 The department provides about 100 courses, of which most are included in the Master's Programs "Biomedical Engineering", "Electric Power Engineering", "Systems, Control and Mechatronics" and "Communication Engineering".



 Read more at www.chalmers.se/en/departments/e2



	Information about the research group

The Computer Vision Group conducts research in the field of automatic image interpretation and perceptual scene understanding. The group targets both medical applications, such as the development of new and more effective methods and systems for analysis, support and diagnostics, as well as general computer vision applications including autonomously guided vehicles (particularly self-driving cars), image-based localization, structure-from-motion and object recognition. The main research problems include mathematical theory, algorithms and machine learning (deep learning) for inverse problems in artifical intelligence.

Project description

We are interested in representation learning, especially learning features from images and videos. Representation learning is one important aspect of deep learning and machine learning, and good representations of input data are essential for the generalization ability, interpretability and robustness of machine learning methods.

In this project we consider "meaningful" representations in the broadest sense as good representations. Meaningful representations can be achieved in several ways, and the goal of this project is to investigate in the following directions in particular: first, constraining the neural network architecture to loosely imitate hand-crafted and therefore interpretable methods; and second, extracting features that allow a fine-grained prediction (such as predicting the next frame in a video or how a scene appears at a novel view-point).

The majority of images used in machine learning applications are captured using a pinhole camera, i.e. images are 2D projections of a 3D world. Hence, imaged objects have an apparent size different from their true size, and objects are subject to full or partial occlusions. The overwhelming majority of deep learning approaches processes images solely as 2D array of pixels, and training data has to cover e.g. objects at different apparent sizes and with a variety of realistic occlusion patterns.

One of the main advantages of having a good representation of input data is the ability to train a machine learning method rapidly, i.e. from a small number of training examples, to solve a new task. Contrary to biological learning systems artificial neural networks require large training sets and many rounds of training to achieve satisfactory performance. The aim of this project is to reduce the required training data in two ways: first, insert knowledge about the relationship between 2D images and the 3D world into deep learning architectures; and second, use cheaply available unsupervised and weakly supervised data to teach such 3D "aware" neural networks the specific structure and characteristics of the actual 3D environment. The features extracted from such a trained networks are expected to be versatile in a number of applications related to understanding of natural images.

Funding has been obtained from the Wallenberg AI, Autonomous Systems and Software Program (WASP) which is Sweden's largest ever individual research program, a major national initiative for strategically basic research, education and faculty recruitment. The program is initiated and generously funded by the Knut and Alice Wallenberg Foundation (KAW) with 2.6 billion SEK. In addition to this, the program receives support from collaborating industry and from participating universities to form a total budget of 3.5 billion SEK. Major goals are more than 50 new professors and more than 300 new PhDs within AI, Autonomous Systems and Software. The vision of WASP is excellent research and competence in artificial intelligence, autonomous systems and software for the benefit of Swedish industry. For more information about the research and other activities conducted within WASP please visit: http://wasp-sweden.org/

Major responsibilities

Your major responsibilities are to pursue your own doctoral studies. You are expected to develop your own scientific concepts and communicate the results of your research verbally and in writing, both in Swedish and in English. The position generally also includes teaching on Chalmers’ undergraduate level or performing other duties corresponding to 20 per cent of working hours.

Full-time temporary employment.The position is limited to a maximum of five years.

Qualifications

To qualify as a PhD student, you must have a master’s level degree corresponding to at least 240 higher education credits in a relevant field (physics, mathematics or computer science).

The position requires sound verbal and written communication skills in Swedish and English. If Swedish is not your native language, you should be able to teach in Swedish after two years. Chalmers offers Swedish courses.

Chalmers offers a cultivating and inspiring working environment in the dynamic city ofGothenburg.

Read more aboutworking at Chalmersand ourbenefitsfor employees.

Application procedure

The application should be marked with Ref 20180717 and written in English.

CV: (Please name the document: CV, Family name, Ref. number)

CV

Other, for example previous employments or leadership qualifications and positions of trust.

Two references that we can contact.

Personal letter: (Please name the document as: Personal letter, Family name, Ref. number)

1-3 pages where you introduce yourself and present your qualifications.

Previous research fields and main research results.

Future goals and research focus.Are there any specific projects and research issues you are primarily interested in?

Copies of bachelor and/or master's thesis.

Attested copies and transcripts of completed education, grades and other certificates, eg. TOEFL test results.

Application deadline: 15 January, 2019

For instructions on how to apply, please refer to: http://bit.ly/2Cin7Ak