Applications are invited for two University Lectureships (equivalent to Assistant Professorships) in the broad areas of Machine Learning or Computer Vision at the University of Cambridge. The successful candidate will join the Information Engineering Division which includes the Computational and Biological Learning Laboratory and the Machine Intelligence Laboratory. 

The candidate will lead a research programme in one or more of the following areas: Computer Vision, Machine Learning and Decision Making.

We encourage applicants who will strengthen our current research activities in probabilistic machine learning, reinforcement learning, supervised and unsupervised learning and computer vision.

These positions have been funded in part by a generous contribution from Toyota Motor Corporation. The successful applicants will have the opportunity and resources to work with Toyota on rich real-world data modelling, prediction, and decision-making problems. We will give priority to candidates who are well placed to do this.

Full advert: http://www.jobs.cam.ac.uk/job/23622/

Deadline: 1 March 2020

The Personalised Medicine research group of the sitem Center for Translational Medicine and Biomedical Entrepreneurship at the University of Bern is offering a PhD position in the research and development of artificial intelligence (AI) algorithms for anatomy modelling and anatomical analysis. The PhD candidate will work within our multinational team and in close collaboration with our industrial, research and clinical partners, towards a cutting-edge software system for the diagnosis, surgical planning and outcome prediction of shoulder pathologies. They will build on deep learningbased algorithms to investigate clinically applicable segmentation solutions for the purpose of tissue composition and morphological analyses, patient specific biomechanical surgical outcome simulations and radiomics-based analysis of treatment outcome success. The student will work within a highly motivated team of experts at the sitem Center and interface with imaging physicists, surgeons and neurologists from the University Hospital of Bern, biomechanical engineers from ETH Zurich and industrial partners.

Please find a more detailed information in the attached pdf.

How to apply
Please send an application letter and CV addressed to Dr. Kate Gerber (kate.gerber@sitem.unibe.ch).
Applications may be considered until the position is filled. The University of Bern is an equal opportunity employer. Qualified female candidates are especially encouraged to apply.

The young Machine Learning Research Laboratory chaired by Prof. Marcus Liwicki has several open positions in the area of Machine Learning and AI. We can offer well-equipped laboratory facilities for performing research and good academic network in Sweden and abroad. The Post-Doc Scholarship is 100% for the duration of at least one year (extendable to two years). Afterwards, hiring at LTU is also possible. The awarded candidate will get a direct stipend of 28 000 SEK per month, which roughly corresponds to 2800 EUR and is 30% above the average net income in the area (more info https://www.numbeo.com/cost-of-living/in/Lulea).

Electroencephalography (EEG) is a key diagnostic tool in epilepsy. The current gold-standard is visual analysis of EEG recordings by trained experts, who look for specific grapho-elements like spike-and-wave discharges or focal slowing. Since spikes (i.e. highly synchronized discharges that are isolated from the background EEG) are considered an abnormal electrical phenomenon associated with epilepsy, they are of special interest in EEG analysis. However, there is conflicting evidence regarding the relationship between spiking and ictogenesis, i.e. the generation of a seizure.

To support diagnostics, in the last 20 years a variety of quantitative EEG (qEEG) analysis methods has been developed. They aim at revealing and quantifying signal properties that may escape expert analysis. However, it has not yet been sufficiently studied, whether qEEG measures are sensitive to phenomena that are independent or complementary to epileptic spikes. In order to approach this question, we need an algorithm to remove spikes from EEG recordings. Analysis of despiked EEG can then be compared to analysis of the original signals.

Jmail et al. (2017) have introduced a spike fitting algorithm that requires spike detection as a first step and makes strong assumptions about the shape of a spike. In the proposed master thesis project we aim at using Generative Adversarial Networks (GANs, Goodfellow et al. 2014) for a similar purpose. Clinical EEGs recorded at the Inselspital (~100 patients, 30-120 channels, sampling rate 512 or 1024 Hz) are available and can be used immediately. Restricting the training data to spike-free EEG, this will allow to generate artificial signals that resemble the original EEG closely in spike-free situations and continuously replaces spikes with plausible EEG-like signals (Hartmann et al. 2018). As a starting point for this GAN-based despiking strategy we will use GAN-based denoising algorithms for physiological signals (Casas et al. 2018, Gandhi et al. 2018).

More information can be found here.

Contact

Christian Rummel (PhD)
christian.rummel@insel.ch
Support Center for Advanced Neuroimaging (SCAN)
Institute of Diagnostic and Interventional Neuroradiology
Inselspital, 3010 Bern

Our team focuses on computational imaging, computational photography, and computer vision with the goal of impacting visual expression and communication. The team has 5 full-time researchers and 3 part-time contractors. Mohit Gupta at UW and Changxi Zheng at Columbia are currently our faculty collaborators.

Openings: We have 3 full-time positions and 6 summer internships open at the moment. The full-time positions are: Research Scientist (recent Phd), Senior Research Scientist (3-10 years since Phd) and Research Software Engineer (with IOS experience).

Strong candidates can send their CVs to Rachel Greenfield (rach@snap.com).