Base Styles/Icons/lock/open Created with Sketch.

AScI internships: Available positions

These topics are offered in the AScI international internship programme 2019.

Positions for summer 2019

1) Field of study: Materials physics, Metallurgy

School / Department: School of Science, Department of Applied Physics

Professor in charge of the topic: Filip Tuomisto

Academic contact person for further information on topic: Filip Tuomisto, [email protected]

Title of topic: Vacancy defects and atomic-scale damage in High Entropy Alloys

Short task description:                     

Materials used in current or future energy production often face physical and chemical environments that are extremely hostile. Combinations of chemically corrosive environments, large heat loads, hard ionizing radiation fluxes and, for example, simultaneous mechanical loads, significantly shorten the lifetime of components manufactured from currently available materials

This project will focus on a novel class of single-phase solid solution metal alloys, called high entropy alloys (HEAs) when the number of alloying elements with significant molar fractions is more than five. This multi-metal approach clearly demonstrates that a breakthrough with traditional metallurgy is possible. Beyond that concept, all the possible strategies planned to improve materials and associated microstructures by design approaches are key factors to develop new classes of (functional or structural) materials with enhanced properties. The design challenges of primary interest are corrosion and embrittlement by hydrogen and other mobile impurities into the original material. HEA performance in corrosive environments is a hot topic globally, while the fundamental understanding of these types of materials from the atomic to macroscale is currently missing from the wider scientific literature.

As an AScI Summer Intern in the Antimatter and Nuclear Engineering group at the Department of Applied Physics, you would quantify and characterize defects in high-entropy alloys using positron annihilation spectroscopy methods. The ideal applicant will be highly motivated, willing to learn, have a basic understanding of materials physics, and fluency in both written and spoken English. Previous experience in the area of condensed matter physics or experimental laboratory work would be considered an added advantage, but is not obligatory.

2) Field of study: Physical chemistry/Electrochemistry

School / Department: School of Chemical Engineering, Department of Chemistry and Materials Science

Professor in charge of topic: Lasse Murtomäki

Academic contact person for further information on topic: Lasse Murtomäki, [email protected]

Title of topic: Redox reactions at the immiscible water-oil interface

Short task description:

A liquid-liquid interface provides a platform to mimic redox reactions in nature where potential differences across interfaces are not controlled by electrodes or auxiliary power sources but by partitioning of ions. Yet, instrumentation can be used in the study of the reactions. In the proposed project, reduction of some metals at the water-oil interface is studied with chemical control of the interfacial potential difference. The key instrument is the Scanning Electrochemical Microscope (SECM) but also other voltammetric measurements will be carried out. The target is in metals that cannot be reduced from aqueous solutions due to their high reduction potentials.

An applicant should be familiar with fundamental physical chemistry (a relevant university course will do), and knowing electrochemistry is a plus. Independent working skills are essential, as in July quite a number of staff take their holidays.

3) Field of study: Soft and active matter physics

School / Department: School of Science, Department of Applied Physics

Professor in charge of topic: Jaakko Timonen

Academic contact person for further information on topic: Jaakko Timonen, [email protected]

Title of topic: Summer internship in experimental soft and active matter physics

Short task description:

Active Matter group carries out experimental research in the field of soft and active matter physics. Systems of interest include colloidal matter, self-propulsive particles, micro-organisms, cellular aggregates, magnetic tweezers, capillary phenomena and ferrofluids. We are looking forward to hiring up to two highly motivated students for summer 2019. The topic(s) for the summer internship(s) will be decided after the interviews based on the skills and interests of the student(s). Mandatory requirements for obtaining an internship position include excellent command of spoken and written English and excellent understanding of foundations of soft matter physics, statistical physics and/or biophysics. Prospective students are encouraged to contact prof. Jaakko Timonen ([email protected]) for details (please attach your CV in the email).

4) Field of study: Physics

School / Department: School of Science, Department of Applied Physics

Professor in charge of topic: Peter Liljeroth

Academic contact person for further information on topic: Peter Liljeroth, [email protected]

Title of topic: Automated atomic assembly using low-temperature scanning tunneling microscope (STM)

Short task description:
The project involves coding computer routines for controlling a low-temperature scanning tunneling microscope (STM) to build atomically well-defined structures on surface atom-by-atom. Subsequently, these routines will be used in the laboratory. The tip of the STM can be used to push/pull individual atoms into the desired positions with atomic-scale precision. The goal of the project is to automatize this process using scripts written in Python, Matlab, or Labview that autonomously scan over the area of interest, identify the positions of the adsorbates that will be manipulated, design suitable paths for the manipulation and carry it out, while monitoring the success of the process in-situ.
Required skills: basic physics courses and programming skills in Python and Matlab. Knowledge of Labview programming a plus.

5) Field of study: Radio Science, Electronics, Biomedicine

School / Department: School of Electrical Engineering, Department of Electronics and Nanoengineering

Professor in charge of topic: Konsta Simovski

Academic contact person for further information on topic: Konsta Simovski, [email protected]

Title of topic: Metasurface-enhanced body coil for high field magnetic resonance imaging of human body

Short task description:

A body coil for high-field MRI of human body performed as a modified birdcage antenna was suggested recently and its high potential for the improvement of the image compared to a conventional birdcage was proved. However, the improvement is achieved either in specific absorption rate (reduced) or signal-to-noise ratio (enhanced). Meanwhile, there is an optimal impedance of the bore internal surface that would allow a combination of high SNR and low SAR. We have an idea how to do it with an explicit metasurface. To check the idea in full-wave simulations and perhaps in a model experiment the applicant should sufficiently know the applied electromagnetics, be capable to apply commercial electromagnetic solvers, be familiar with Matlab (or Matematica) and enthusiastic in research, in general. Understanding of the underlying physics and experimental skills are strongly desirable.

6) Field of study: Computer Science

School / Department: School of Science, Department of Computer Science

Professor in charge of topic: Stavros Tripakis

Academic contact person for further information on topic:

Title of topic: Checking approximate equivalence of neural networks using SMT solvers

Short task description:

The candidate will study equivalence of neural networks. After selecting a sufficiently simple initial example (for understandability), and defining some notions of equivalence, including approximate equivalence using well-defined metrics, the candidate will devise methods to check such equivalence. Possible ways of checking equivalence would be to encode the problem as a satisfiability problem in a given theory and using existing SMT solvers to check it.

7) Field of study: Computer science (Machine learning)

School / Department: School of Science, Department of Computer Science

Professor in charge: Pekka Marttinen

Contact: [email protected]

Title: Machine Learning for Health

Abstract:

Recent years have witnessed accumulation of massive amounts of health related data, enabling researchers to address problems such as: how to allocate health care resources fairly and efficiently, how to provide personalized guidance and treatment to a user based on time-series data from wearable self-monitoring devices, or how to use genomic data to understand disease or antibiotic resistance. Answering these questions requires new machine learning methodology to be developed. We have several interdisciplinary research projects ongoing, where the goal is to design new machine learning models and algorithms for applications in health and welfare, together with leading experts in the respective fields from Finland and abroad. Examples of applications include: analysis of electronic health records, mobile health, genomics, antibiotic resistance, epidemiology. Successful applicants for a summer internship are expected to have an outstanding record in computer science, mathematics, statistics, or a related field, and a strong interest to focus on one or more of the following topics: probabilistic machine learning, deep learning, bioinformatics, Bayesian modeling, causal learning, time-series modeling, likelihood-free inference.

8) Field of study: Theoretical physics (quantum computing)

School / Department: School of Science, Department of Applied Physics
Professor in charge of topic: Prof. Christian Flindt
Academic contact person for further information on topic: Aydin Deger, [email protected]
Title of topic: Implementation of quantum algorithms on quantum computers
Short task description:
The applicant should have a solid understanding of basic quantum mechanics and statistical physics. Experience with numerical calculations is welcome but not required.

9) Field of study: Theoretical physics (phase transitions)

School / Department: School of Science, Department of Applied Physics
Professor in charge of topic: Prof. Christian Flindt
Academic contact person for further information on topic: Aydin Deger, [email protected]
Title of topic: Lee-Yang zeros and large-deviation statistics at phase transitions
Short task description:
The applicant should have a solid understanding of basic quantum mechanics and statistical physics. Experience with numerical calculations is welcome but not required.

10) Field of study: Theoretical physics (quantum thermodynamics)

School / Department: School of Science, Department of Applied Physics
Professor in charge of topic: Prof. Christian Flindt
Academic contact person for further information on topic: Paul Menczel, [email protected]
Title of topic: Maximizing thermodynamic precision in open quantum systems
Short task description:
The applicant should have a solid understanding of basic quantum mechanics and statistical physics. Experience with numerical calculations is welcome but not required.

11) Field of study: Theoretical physics (coherent single-electron transport)

School / Department: School of Science, Department of Applied Physics
Professor in charge of topic: Prof. Christian Flindt
Academic contact person for further information on topic: Elina Potanina, [email protected]
Title of topic: Coherent single-electron transport in periodically driven systems
Short task description:
The applicant should have a solid understanding of basic quantum mechanics and statistical physics. Experience with numerical calculations is welcome but not required.

12) Field of study: Computer Science

School / Department: School of Science, Department of Computer Science

Professor in charge of topic: Juho Rousu

Academic contact person for further information on topic: Juho Rousu, [email protected]

Title of topic: Predicting structured output with deep kernel regression models

Short task description:

Structured output prediction is a supervised machine learning problem where both input and output variables are complex structured objects such as sequences, trees or graphs. One approach to structured output prediction consist in two steps: first solve a regression problem in a well-chosen output feature space and then solve a pre-image problem to get a solution in the original output space. Recently, this kind of approach has been extended to deep learning, allowing the methods to model complex non-linear dependencies and potentially higher predictive accuracy.

The task of the intern is to study the performance of novel deep kernel regression models on datasets arising from biomedicine, in particular small molecule identification, where regression based structured output methods currently are state-of-the-art (Brouard, et al. 2016). The project is a part of an on-going research collaboration between KEPACO group at Aalto (Prof. Juho Rousu) and Telecom Paristech (Prof. Florence d’Alche-Buc). The internship gives good possibilities for continuing as a thesis project.

Prerequisite skills: calculus, linear algebra and basic probability theory, algorithms, programming skills, basics in machine learning and/or data science.

References:

Brouard, C., Shen, H., Dührkop, K., d'Alché-Buc, F., Böcker, S. and Rousu, J., 2016. Fast metabolite identification with input output kernel regression. Bioinformatics32(12), pp.i28-i36.

13) Field of study: Computational physics

School / Department: School of Science, Department of Applied Physics

Professor in charge of topic: Filip Tuomisto

Academic contact person for further information on topic: Academy Research Fellow Ilja Makkonen, [email protected]

Title of topic: Computational study of positron states and annihilation in solids

Short task description:
Our group uses the positron annihilation technique for experimental detection, quantification and characterization of point defects in solids and also develops state-of-the-art computational techniques for supporting the experimental activities. The following review gives some idea of the kind of work done within this area: "Defect identification in semiconductors with positron annihilation: Experiment and theory", Reviews of Modern Physics 85, 1583 (2013).

The available computational materials and positron physics projects involve application and/or development of atomistic density-functional or quantum many-body (quantum Monte Carlo) simulation techniques for positron-defect interaction in solids.

Prospective candidates should possess at least a basic knowledge on quantum mechanics. Good programming skills are a big asset.

14) Field of study: Electrical engineering, Mechatronics, robotics and automation, Measurement technology, Engineering physics

School / Department: School of Electrical Engineering, Department of Electrical Engineering and Automation

Professor in charge of topic: Quan Zhou

Academic contact person for further information on topic: Quan Zhou, [email protected]

Title of topic: Summer Trainee in Robotic Instruments

Short task description:

Do you want to develop, program and test intelligent, miniature robotic systems that can analyze and treat biological cells, producing biomaterials, or measuring nanoscopic properties of new materials?  We are looking for summer trainee who is willing to explore this highly interdisciplinary robotic instruments field. The actual task will be related to one or several of our ongoing research projects and can be decided based on mutual interests of the applicant and the research group. The position is available for summer 2019 (June-August or agreed separately).

REQUIREMENTS AND DETAILS

As a successful applicant you should be highly motivated in learning interdisciplinary skills and eager to work hands-on. The student should have education in electrical engineering, mechatronics, robotics and automation, measurement technology, engineering physics, or related fields.

15) Field of study: Computer Science

School / Department: School of Science, Department of Computer Science

Professor in charge of topic: Mario Di Francesco

Academic contact person for further information on topic: Mario Di Francesco, [email protected]

Title of topic: Mobile Computing and the Internet of Things

This summer internship is related to current research on mobile computing and the Internet of Things in Mario Di Francesco’s group (https://users.aalto.fi/difram1/). The student involved in the summer inter will conduct a research and development task in one of the following topics: camera-display communications; edge and fog computing; architectures and protocols for the Internet of Things (IoT); performance of distributed applications. Multiple positions are available.

16) Field of study: Machine Learning

School / Department: School of Science, Department of Computer Science

Professor in charge of topic: Prof. Samuel Kaski, [email protected]

Academic contact person for further information on topic (name & email): Prof. Samuel Kaski, [email protected]

Title of topic: Approximate Bayesian Computation

Short task description:

For this project you will join the Aalto Probabilistic Machine Learning Group (http://research.cs.aalto.fi/pml/) to develop probabilistic models and inference techniques. We are developing Approximate Bayesian Computation (ABC) techniques for inference in simulator-based models. The work is basic research in machine learning and inference, with applications chosen to match your interests. We have particularly exciting work on-going in modeling of human behavior and precision medicine. For more information see our ABC software ELFI http://elfi.readthedocs.io and for instance https://arxiv.org/abs/1708.00707, https://academic.oup.com/sysbio/article/66/1/e66/2420817 . Students with strong background in mathematics and interest in model development are especially encouraged to apply. Skills and interest in programming are a big plus.

17) Field of study: Machine Learning

School / Department: School of Science, Department of Computer Science

Professor in charge of topic: Prof. Samuel Kaski, [email protected]

Academic contact person for further information on topic: Prof. Samuel Kaski, [email protected]

Title of topic: Human-in-the-loop machine learning and human-AI collaboration

Short task description:

Humans are increasingly interacting with machine learning based adaptive systems, both in industrial settings and as end-user consumers. How to optimally combine the strengths of humans and machines is one of the most interesting scientific questions at the moment. We are developing new approaches and applications for interactive human-in-the-loop machine learning and human-AI collaboration, with the aim of increasing the performance and efficiency of the systems and for improving the user experience. This project lies at the intersection of machine learning, human-computer interaction, and cognitive science. The summer intern will work as a member in a project involving one or more of the following machine learning methodologies, depending on the open research questions in the project and the intern's interests: reinforcement learning, inference in simulator based models, probabilistic modelling and  programming, and deep learning. Position requires programming skills in Python.

Link: http://research.cs.aalto.fi/pml

18) Field of study: Materials science, engineering physics

School / Department: School of Electrical Engineering, Department of Electrical Engineering and Automation & School of Science, Department of Applied Physics

Professors in charge of topic: Quan Zhou & Robin Ras

Academic contact persons for further information on topic: Matti Hokkanen, [email protected] & Maja Vuckovac, [email protected]

Title of topic: Scanning Droplet Adhesion Microscopy: Development of Application Cases

Short task description:

Scanning droplet adhesion microscopy (SDAM) technique, developed jointly in the research groups of Prof. Quan Zhou (https://www.aalto.fi/department-of-electrical-engineering-and-automation/robotic-instruments ) and Prof. Robin Ras (http://physics.aalto.fi/smw ) in Aalto University, enables quantitative assessment of the wettability of various liquid-repellent surfaces in unprecedented detail [1]. In this project, you have a chance to join an international, highly ambitious team on a quest to realize the full potential of the technique by demonstrating new measurement applications, both scientific and industrial. 

Your initial role is to support scientists and engineers in the development of the technique by engaging in tasks such as sample preparation, measurement calibration and analysis of measurement data. Later, you will get a chance to conduct independent measurements on various samples of interest, while also serving as a test user in usability trials for a new proof-of-concept instrument. 

The project is suitable for a highly motivated student with interest in practical measurement and laboratory work and background in physical sciences, chemistry, materials technology or engineering. You will develop skills and understanding in e.g. surface engineering, materials science, measurement technique, data analysis and project management. The work is connected to the Business Finland TUTL project Scanning Droplet Adhesion Microscopy.

[1] V. Liimatainen, M. Vuckovac, V. Jokinen, V. Sariola, M. J. Hokkanen, Q. Zhou & R. H. A. Ras, Mapping microscale wetting variations on biological and synthetic water-repellent surfaces, Nature Communications 8, 1798 (2017). http://dx.doi.org/10.1038/s41467-017-01510-7

19) Field of study: Physics, materials science, or engineering

School / Department: School of Science, Department of Applied Physics

Professor in charge of topic: Robin Ras

Academic contact person for further information on topic:  Dr Matilda Backholm, [email protected], backholm.wordpress.com

Title of topic: Friction of drops moving on superhydrophobic surfaces

Short task description:

Superhydrophobicity, that is the extreme fear of water, is a fascinating surface property found in nature on many plants and insects. As is beautifully exemplified by a lotus leaf, a water drop on a superhydrophobic substrate will be almost spherical in shape. Due to the incredibly poor wetting properties of the surface, the water drop can easily roll off with a tiny resistance. By mimicking the design found in nature, scientists have manufactured artificial superhydrophobic surfaces with self-cleaning, non-wetting, anti-icing, and anti-fogging properties, just to name a few examples. Measuring the friction force of drops moving on these extremely slippery substrates is important to further improve their quality, as well as to understand the physics governing the drop motion.

In this research project, you will perform cutting-edge micromechanical friction measurements on artificial superhydrophobic surfaces using an existing setup in our lab. You will learn how to perform the experiments and how to analyze the results with MATLAB. This summer research assignment is suitable for a student with a genuine interest in performing experimental research in a dynamic, interdisciplinary environment. A background in physics, materials science, or engineering will be beneficial.

20) Field of study: Computer Science / Computational Biology

School / Department: School of Science, Department of Computer Science

Professor in charge of topic: Pekka Orponen

Academic contact person for further information on topic: Pekka Orponen, [email protected]

Title of topic: Algorithms for the design of RNA nanostructures

Short task description:

The area of DNA nanotechnology [1] employs DNA as generic building material for assembling nanoscale objects with dimensions in the order of 10-100 nanometres. For instance, our group recently demonstrated, together with a biochemistry team from Karolinska Institutet, a general technique for rendering almost arbitrary 3D wireframe designs into biomolecules folded from a single long DNA strand [2].

For several reasons, there is increasing interest in the DNA nanotechnology community to move from DNA to RNA as source material. This is however very challenging, because RNA has remarkably much richer and less well understood folding kinetics than DNA. Thus a current focus of our group is to develop algorithmic methods and tools to support the task of designing complex 2D and 3D RNA nanostructures. Some preliminary results of our work are described in the summary [3].

The topic of this internship project is to learn about the present combinatorial models of RNA folding and develop algorithmic methods for designing RNA sequences that fold into desired 2D or 3D shapes. Simulation studies will then be used to screen the proposed designs, towards a possibility of eventual validation by laboratory experiments.

The project requires familiarity with basic algorithm design techniques, facility with combinatorial thinking, and good programming skills. Previous knowledge of biomolecules is not necessary, although it is an asset. The work is performed in the context of research project “Algorithmic designs for biomolecular nanostructures (ALBION)”, funded by the Academy of Finland. For further information, please see the research group webpage at http://research.cs.aalto.fi/nc/ .

Online materials:

[1] https://en.wikipedia.org/wiki/DNA_nanotechnology

[2] http://old.cs.aalto.fi/en/current/news/2015-07-23/

[3] http://research.cs.aalto.fi/nc/sources/EKOM18.pdf

21) Field of study: Computer Science / Computational Biology

School / Department: School of Science, Department of Computer Science

Professor in charge of topic: Pekka Orponen

Academic contact person for further information on topic: Dr. Vinay Gautam, [email protected]

Title of topic: Simulation of kinetically-controlled DSD systems

Short task description:

One of the goals of DNA nanotechnology is to design enzyme-free DNA-based molecular systems with programmable dynamic behaviours. The invention of the DNA Strand Displacement (DSD) reaction [1] provides a basic building block which can be cascaded in a variety of ways to implement DNA-based reaction networks. In general, Chemical Reaction Networks (CRN) can be used to describe arbitrary dynamic behaviours, which can be translated into a set of DSD reactions.

Although a variety of DSD systems have been demonstrated both theoretically and experimentally, one of the main bottlenecks that has limited the scale of DSD systems is the undesired triggering of displacement reactions causing leakages at various levels, which result in unreliable and unpredictable behaviours.

Our on-going project is investigating an alternative implementation of CRNs using kinetically-controlled robust designs of DSD reactions.  The task of this internship project is design a dynamic simulation framework, something similar to Microsoft’s Visual DSD [2] with a very basic set of functionalities based on ordinary differential equations of the modified DSD reaction systems.

The project requires good programming skills and some understanding of ordinary differential equations. Previous knowledge of biomolecules and stochastic simulation is not necessary, although it is an asset. The work is performed in the context of research project “Algorithmic designs for biomolecular nanostructures (ALBION)”, funded by the Academy of Finland. For further information, please see the research group webpage at http://research.cs.aalto.fi/nc/ .

Online sources:

[1] https://en.wikipedia.org/wiki/Toehold_mediated_strand_displacement

[2] https://bit.ly/2Gy8ehv

22) Field of study: Computer Science / Computational Biology

School / Department: School of Science, Department of Computer Science

Professor in charge of topic: Pekka Orponen

Academic contact person for further information on topic: Dr Vo Hong Thanh, [email protected]

Title of topic: Stochastic simulation for computational biology and biotechnology

Short task description:

The topic of this internship project is to learn about the stochastic approach for stochastic processes with applications to computational biology and biotechnology. Biological processes at the nanoscale are inherently stochastic. Molecules driven by random walk move around in a solution (for instance, the inside of a cell or a test tube) and experience a series of collisions with each other.  A collision between molecular species forms a reaction if it satisfies specific reaction conditions, e.g., activation energy. The result of a reaction can be the forming of new chemical bonds, breaking existing ones or sometimes both and ultimately leading to the production of necessary molecular species to help perform the activities of the biological system. Stochastic processes provide a powerful framework to study the stochastic nature of biological systems. Each biological event is modelled as a stochastic process and the temporal dynamics of the system is then studied by the stochastic simulation technique (https://en.wikipedia.org/wiki/Stochastic_simulation).

The project aims to develop computational methods for simulating and analysing stochastic dynamics of biological processes such as regulatory biochemical reactions or the folding of RNA sequences. The project requires familiarity with basic probability and algorithm design techniques, together with good programming skills. Previous knowledge of biochemistry is not necessary, although it is an asset. The work is performed in the context of research project “Algorithmic designs for biomolecular nanostructures (ALBION)”, funded by the Academy of Finland. For further information, please see the research group webpage at http://research.cs.aalto.fi/nc/ .

23) Field of study: Computer science, in particular machine learning and human-computer interaction

School / Department: School of Electrical Engineering, Department of Communications and Networking

Institute: Finnish Center for AI (fcai.fi)

Professor in charge of topic:  Associate Professor Antti Oulasvirta

Academic contact person for further information on topic: Antti Oulasvirta, [email protected]

Title of topic: Design mining

Short task description:

The intern will assist in the study and implementation of a data-driven method for computational design. Design mining refers to modeling of design features in a domain (e.g., web pages in social media). A large scale dataset is provided, and the task is to build models of elementary features, such as locations and color distribution of elements. This kind of model can be used in the generation of designs and their adaptation, as well as in design tools.

24) Field of study: Computer science, in particular machine learning and human-computer interaction

School / Department: School of Electrical Engineering, Department of Communications and Networking

Institute: Finnish Center for AI (fcai.fi)

Professor in charge of topic:  Associate Professor Antti Oulasvirta

Academic contact person for further information on topic: Jussi Jokinen, [email protected]

Title of topic: Reinforcement learning in complex interactive tasks

Short task description:

The intern will assist in developing a hierarchical reinforcement learning model for simulating interactive adaptive behaviour in complex interactive human-computer interaction task. The general model will be provided, and the intern will modify it to fit to a scenario that is selected together with the supervisor. The scenario will involve a problem solving process, where the agent has multiple, possibly conflicting goals, and the actions taken by the agent have dynamic outcomes. The intern is expected to have some knowledge on reinforcement learning, and programming ability in Python.

25) Field of study: Computer science, in particular human-computer interaction and machine learning

School / Department: School of Electrical Engineering, Department of Communications and Networking

Institute: Finnish Center for AI (fcai.fi)

Professor in charge of topic:  Associate Professor Antti Oulasvirta

Academic contact person for further information on topic: Luis A. Leiva, [email protected]

Title of topic: Addressing uncertainty in 2D gesture input

Short task description:

Gestures represent user interface commands, typically those commands that are very repetitive (e.g. dialing up a family member) or tedious to access (e.g. selecting an option from a multi-level menu). Gesture input relies on gesture recognizers to decide which command should be triggered. Because gesture recognizers are not error-free, unintended or potentially harmful commands could be triggered involuntarily. In such cases, it would be useful to ask the user to disambiguate among perceptually similar gestures. But it is important not to overwhelm the user and only ask as few times and as few information as possible. The intern will assist in the study and implementation of a prototype that aims to answer the following research questions: When should the user be asked for disambiguation? How much information should be presented to the user to disambiguate? The intern is expected to have basic knowledge on statistics and machine learning, and good programming skills in JavaScript and/or Python.

26) Field of study: Computer science, in particular human-computer interaction and machine learning

School / Department: School of Electrical Engineering, Department of Communications and Networking

Institute: Finnish Center for AI (fcai.fi)

Professor in charge of topic:  Associate Professor Antti Oulasvirta

Academic contact person for further information on topic: Luis A. Leiva, [email protected]

Title of topic: Addressing uncertainty in motion gaming

Short task description :

In motion gaming systems, players interact with the system through gestures and body movements. Motion control technologies are also being applied to many non-gaming areas such as TVs, wall-sized displays, and IoT devices, just to name a few. User input relies on gesture recognizers to help the system decide which command should be triggered. Because gesture recognizers are not error-free, unintended or potentially harmful commands could be triggered involuntarily. In such cases, it would be useful to ask the user to disambiguate among perceptually similar gestures. But it is important not to overwhelm the user and only ask as few times and as few information as possible. The intern will assist in the study and implementation of a prototype that aims to answer the following research questions: When should the user be asked for disambiguation? How much information should be presented to the user to disambiguate? The intern is expected to have basic knowledge on statistics and machine learning, and good programming skills in JavaScript and/or Python.

27) Field of study: Computer science, in particular human-computer interaction and machine learning

School / Department: School of Electrical Engineering, Department of Communications and Networking

Institute: Finnish Center for AI (fcai.fi)

Professor in charge of topic:  Associate Professor Antti Oulasvirta

Academic contact person for further information on topic: Luis A. Leiva, [email protected]

Title of topic: Neural gesture synthesis

Short task description:

Training a high-quality gesture recognizer requires providing a large number of examples to enable good performance on unseen, future data. However, recruiting participants, data collection and labeling, etc. necessary for achieving this goal are usually time-consuming and expensive. Thus, it is important to investigate how to empower developers to quickly gather gesture samples. This project aims to solve this by means of synthetic data: given a user-provided example, a computerized system will generate as many variations as possible of such a provided example. The intern will assist in the development of a neural network that implements this programming-by-example principle. The intern is expected to be familiarized with basic machine learning principles (will be trained otherwise) and Python's Tensorflow and Keras frameworks.

28) Field of study: Computer science, in particular human-computer interaction and information retrieval

School / Department: School of Electrical Engineering, Department of Communications and Networking

Institute: Finnish Center for AI (fcai.fi)

Professor in charge of topic:  Associate Professor Antti Oulasvirta

Academic contact person for further information on topic: Luis A. Leiva, [email protected]

Title of topic: Generating user interface mockups

Short task description:

The logical place to start the actual design process of a user interface is with a rough wireframe; i.e. a low-fidelity, simplified outline of the user interface elements. However, wireframes are not very visual as they just provide the skeleton of those elements. Mockups, on the contrary, allow designers to solidify their visual decisions and experiment with variations. The problem is that the design space is massive and so it is unclear how to assist designers with appealing mockups in an automated way. The goal of this project is creating a tool to transform low-fidelity wireframes into high-fidelity mockups, in an interactive setting. First, the tool will allow users to draft the user interface via drag and drop elements. Next, the tool will retrieve previously designed elements from a large pool of user interface mockups to "fill in the gaps". Getting the right look and feel (i.e. selecting the right combination of retrieved elements) is an open problem, so there is ample room for creativity here. The intern will assist in the development of said tool, especially in the retrieval part. The intern is expected to have good programming skills in JavaScript or Python.

29) Field of study: Computer science, in particular human-computer interaction and software engineering

School / Department: School of Electrical Engineering, Department of Communications and Networking

Institute: Finnish Center for AI (fcai.fi)

Professor in charge of topic:  Associate Professor Antti Oulasvirta

Academic contact person for further information on topic: Luis A. Leiva, [email protected]

Title of topic: Generating user interface manuals of web applications

Short task description:

Documenting user interfaces is an important but tedious process. It is particularly tedious for web applications, since they can change slightly over time, especially with regard to the visual properties of the interface elements, and so the associated documentation can become outdated quickly. This project aims to solve this problem by creating a prototyping tool that generates user interface manuals of web applications in real time. The intern will assist in the development of said tool, as well as on useful derivations thereof such as a templating system or different output formats (e.g. HTML or PDF). The intern is expected to have good programming skills in JavaScript and be familiarized with the DOM (traversing, manipulation, etc.) Notions of web design will be highly appreciated.

30) Field of study: Computer science, in particular, human-computer interaction and usability testing

School / Department: School of Electrical Engineering, Department of Communications and Networking

Institute: Finnish Center for AI (fcai.fi)

Professor in charge of topic:  Associate Professor Antti Oulasvirta

Academic contact person for further information on topic: Sunjun Kim, [email protected]

Title of topic: Usability and Performance Evaluation of Analog Control Input Devices

Short task description:

Emerging technology opens additional sensing capability on existing input devices. For example, a new sensor embedded on the keyboard allows analog control of keypress rather than discrete on/off control (see https://aimpad.com/ for example). Variety of input devices (keyboard, mouse, gamepad, joystick etc.) are subject to be tested. The intern will explore new design possibilities of such the devices, and assist in the evaluation study and implementation of them. The evaluation will measure quantitative performance metrics (e.g., Fitts’ law task and steering law task) and qualitative usability scores. The intern is expected to have good expertise in usability testing in the human-computer interaction field, good programming skills in graphics (processing, Unity, etc. Basically 2D graphics, but expertise in 3D will be appreciated), data processing and statistical testing, and a basic understanding of electrical engineering.

31) Field of study: Mathematical modelling, Combinatorial optimization, Graphical User interface design, Development of code in Python and Java

School / Department: School of Electrical Engineering, Department of Communications and Networking

Institute: Finnish Center for AI (fcai.fi)

Professor in charge of topic:  Associate Professor Antti Oulasvirta

Academic contact person for further information on topic: Niraj Ramesh Dayama, [email protected]

Title of topic: Automated UI Layout system

Short task description:

Researchers at the User Interfaces group in Aalto University have developed efficient mathematical models to assist in the automated design of graphical user interfaces (GUIs). The resulting GUIs are assured to satisfy several important criterion. These mathematical models now require to be translated into a flexible interface for use of the wider academic/research community. This translation of models into a usable interface will require understanding of Python, Java and web technologies. The students will also gain insights in combinatorial optimization techniques including mixed integer linear programming and genetic algorithms.

32) Field of study: Mathematical modelling, Combinatorial optimization, Development of code in Python and Java, machine learning

School / Department: School of Electrical Engineering, Department of Communications and Networking

Institute: Finnish Center for AI (fcai.fi)

Professor in charge of topic:  Associate Professor Antti Oulasvirta

Academic contact person for further information on topic: Niraj Ramesh Dayama, [email protected]

Title of topic: Integration of integer programming techniques with machine learning

Short task description:

This project is an investigation of whether mixed integer linear programming techniques can be used in conjunction of machine learning algorithms to solve classical academic/research problems. While integer programming is a well proven technique for finding optimal solutions in a wide combinatorial scenario, machine learning can assist in classifying candidate solutions based on several criteria that are not amenable to integer programming. The current project intends to utilize machine learning as an intermediate evaluator for an integer programming framework. If this framework can be proven for a classical test scenario, it can open up a new technique for a wide variety of problems in future.

33) Field of study: Computer science, in particular end-user development and human-computer interaction

School / Department: School of Electrical Engineering, Department of Communications and Networking

Institute: Finnish Center for AI (fcai.fi)

Professor in charge of topic:  Associate Professor Antti Oulasvirta

Academic contact person for further information on topic: Markku Laine, [email protected]

Title of topic: Conflict Resolution in Web Applications

Short task description:

End-user development (EUD) of web applications deals with approaches, languages, and tools that aim to empower end users to create, modify, and extend web applications on their own. One such approach is Mavo (https://mavo.io) that allows end-user developers to describe interactive, data-driven web applications using nothing more than HTML. The intern will assist in the study and implementation of a conflict resolution plugin for Mavo. The intern is expected to have good programming skills in JavaScript and basic knowledge on NoSQL databases.

34) Field of study: Computer science, human-computer interaction, biomedical data-analysis and experimentation

School / Department: School of Electrical Engineering, Department of Communications and Networking

Institute: Finnish Center for AI (fcai.fi)

Professor in charge of topic:  Associate Professor Antti Oulasvirta

Academic contact person for further information on topic: Tuukka Ruotsalo, [email protected]

Title of topic: fNIRS neuroimaging data collection and analysis for affective user interfaces

Short task description:

Novel neuroimaging technology enables measurements of human brain functions during human-computer interaction allowing new types of interactive methods and systems that react to human cognition. The intern will assist experimentation and data-analysis of fNIRS neuroimaging data together with the team of researchers. The overall aim is to automatically detect human’s affective states from brain-signal measurements via fNIRS.

35) Field of study: Computer science, in particular human-computer interaction

School / Department: School of Electrical Engineering, Department of Communications and Networking

Institute: Finnish Center for AI (fcai.fi)

Professor in charge of topic:  Associate Professor Antti Oulasvirta

Academic contact person for further information on topic (name & email): Aurélien Nioche, [email protected]

Title of topic: Collaboration between AI and human: How drawing inferences about cognitive features of user might improve it?

Short task description:

The project the intern will take part in aims to improve interaction between human and artificial intelligence (AI), by providing to the AI with tools that allows it to infer cognitive features of the user while interacting. This project is based on the use of human-computer interaction tasks (i.e. scenarios) that serve as a sandbox to assess how including cognitive modeling of the user might improve this interaction. It includes in silico experiments (with artificial agents only) and lab experiments (with human embodied users). The intern will assist in developing cognitive models of the user used by the AI. The outlines of the task, as well as the model at use for the AI, will be provided, but the intern will be invited to modify it in order to be able to test its own hypothesis. The intern is expected to have some knowledge in data analysis, cognitive modeling and/or in one object-oriented programming language (ex: Python, C#, C++, Java, ...).

36) Field of study: Computer science, in particular human-computer interaction

School / Department: School of Electrical Engineering, Department of Communications and Networking

Institute: Finnish Center for AI (fcai.fi)

Professor in charge of topic:  Associate Professor Antti Oulasvirta

Academic contact person for further information on topic: Kashyap Todi, [email protected]

Title of topic: Web-Based Mixed-Initiative Sketching Tool

Short task description:

Sketching is an essential design activity used to create solutions for a given design task. It is used widely by interface designers, while creating layout designs, for interfaces such as websites or apps. In mixed-initiative sketching, a designer collaborates with the system to create good solutions. The system can present suggestions, or make corrections to the design. Sketchplorer

(www.kashyaptodi.com/sketchplore) is one such system that implements an interactive sketching tool that combines a designer and an optimiser.

In this internship, you will build upon this concept, and develop a web-based intelligent sketching tool that assists designers in constructing user interface layouts.

Reference: Kashyap Todi, Daryl Weir, and Antti Oulasvirta. Sketchplore: Sketch and Explore with a Layout Optimiser. In Proceedings of the 2016 ACM Conference on Designing Interactive Systems (DIS '16). DOI: https://doi.org/10.1145/2901790.2901817

Prerequisites: Web programming skills (HTML, Javascript, etc.)

37) Field of study: Computer science, in particular human-computer interaction

School / Department: School of Electrical Engineering / Department of Communications and Networking

Institute: Finnish Center for AI (fcai.fi)

Professor in charge of topic:  Associate Professor Antti Oulasvirta

Academic contact person for further information on topic: Kashyap Todi, [email protected]

Title of topic: Constructing Machine Learning Models Interactively

Short task description:

Constructing an effective model from data is a crucial yet difficult task for non-experts lacking sufficient training in statistics. Additionally, created models often lack explainability or are hard to understand. In this internship, you will build an interactive tool to facilitate the construction of a model for non-experts and novices. The tool will automate some of the laborious steps in the modelling task, and provide users with guidance when required. As a result, users will be able to construct models that are accurate yet interpretable.

Prerequisites: Some experience with statistical modelling or machine learning (e.g. linear regression). Additionally, some GUI programming skills (web or desktop) would be beneficial, although not mandatory.

38) Field of study: Computer science, in particular human-computer interaction

School / Department: School of Electrical Engineering, Department of Communications and Networking

Institute: Finnish Center for AI (fcai.fi)

Professor in charge of topic:  Associate Professor Antti Oulasvirta

Academic contact person for further information on topic: Kashyap Todi, [email protected]

Title of topic: Semi-Automatic Interface Layout Segmentation

Short task description:

A user interface is a composition of several elements, placed on a two-dimensional canvas. These elements can be of different types, such as texts, images, titles, buttons, etc. Given an image of the user interface, segmenting it accurately into different elements automatically is often error-prone or hard to achieve. On the other hand, when a human visually inspects a layout, they can easily detect the different segments (or elements) of the interface. However, specifying this to a computer can be a tedious task. During this internship, you will build a tool that aids humans in segmenting a layout into a set of elements. Using the tool, the user and the machine will collaboratively achieve the task of fully specifying individual elements and groups within an interface layout.

Prerequisites: GUI programming skills (web or desktop)

39) Field of study: Computer science, in particular human-computer interaction

School / Department: School of Electrical Engineering, Department of Communications and Networking

Institute: Finnish Center for AI (fcai.fi)

Professor in charge of topic:  Associate Professor Antti Oulasvirta

Academic contact person for further information on topic: Kashyap Todi, [email protected]

Title of topic: Self-Adapting User Interfaces

Short task description:

Adaptive interfaces change an application’s visual layout and behaviour depending on the user’s context, skills, or abilities. By continually capturing some information from the user, and modelling them, they can modify their appearance or some characteristics over time, thus improving usability. One example for such an interface is a menu that records user’s clicks and actions, and uses this to decide the ordering of items, or highlight certain items. During this internship, you will further explore the area of self-adapting user interfaces. There are several possibilities for the exact goal of the adaptive user interface, and this can be discussed at the beginning of the internship.

Prerequisites: GUI programming skills (web or desktop)

40) Field of study: Computer science and human-computer interaction

School / Department: School of Electrical Engineering, Department of Communications and Networking

Institute: Finnish Center for AI (fcai.fi)

Professor in charge of topic:  Associate Professor Antti Oulasvirta

Academic contact person for further information on topic: Janin Koch, [email protected]

Title of topic: Interactive Machine Learning in Design Practice

Short task description:

Creative AI for inspiration presents a new challenge for ML and HCI. It requires new ways to interact and express ideas from both the system and its users.

This work is based on an earlier prototype combining interactive machine learning (contextual bandits) and computer vision to assist designers in creating visual collages. In the next step we intend to deploy this system in a long term study with professional designers. The intern will assist in developing, testing and deploying a web based interactive machine learning system for design practice. The original prototype is implemented using python, javascript and css. The intern is expected to have knowledge in Python and web GUI programming.

41) Field of study: Physics

School / Department: School of Science, Department of Applied Physics

Professor in charge of topic: Mikko Alava

Academic contact person for further information on topic: Mikko Alava, [email protected]

Title of topic: Magnetic foams

Short task description:

Here we study complex systems in the shape of magnetic foams. These are made of foams where magnetic nanoparticles have been mixed in. The intern is supposed to be interested in the physics of soft matter, and have skills or be interested in learning experimental physics and data analysis. The project is a collaboration between two groups at the department (Alava and ERC Grant winner Jaakko Timonen) and involves experimental studies of such foams.

42) Field of study: Physics

School / Department: School of Science, Department of Applied Physics

Professor in charge of topic: Mikko Alava

Academic contact person for further information on topic: Antti Puisto, [email protected]

Title of topic: Foam Coarsening

Short task description:

Foams coarsen or change their structure with time. This is a complex process where the microscopical processes vary and the topology of the foam changes with time. Here, we attack this problem by 3D simulations of foams using a novel GPU- based simulation code. The candidates should have an interest in soft matter physics and/or computational physics.

43) Field of study: Computer science

School / Department: School of Science, Depart of Computer Science

Professor in charge of topic: Hong-Linh Truong

Academic contact person for further information on topic: Hong-Linh Truong, [email protected]

Title of topic: Performance and Data Quality Monitoring in IoT Big Data Systems

Short task description:

The internship topic will focus on the development of a system for monitoring and analysis of performance of IoT big data systems. In particular, the work will focus on the following aspects:

  • Leveraging existing monitoring systems like Fluentd, Prometheus, and ElasticSearch to work on monitoring probes for IoT big data systems
  • Establishing knowledge base about IoT big data system components, types of data and infrastructures
  • Providing big NoSQL database for managing the above-mentioned monitoring and knowledge

Strong skills in native cloud programming and big data with strong understanding about performance and monitoring are required. The internship should be able to work with various open source software for big data systems (e.g., Hadoop, Spark, and MQTT) and public cloud systems like Google Cloud and have good knowledge about DevOps (e.g., CI/CD with Jenkins) and Infrastructure-as-Code (e.g, such as Docker containers and Kubernetes).

44) Field of study: Machine learning / computational biology

School / Department: School of Science, Department of Computer Science

Professor in charge of topic: Harri Lähdesmäki

Academic contact person for further information on topic: Harri Lähdesmäki, [email protected]

Title of topic: Bayesian deep learning for personalised medicine

Short task description:

We are looking for several summer internship students to work on probabilistic machine learning and deep generative models for biomedical and health applications. Our on-going research projects involve several important clinical challenges, such as (i) personalized prediction of immunotherapy efficiency for cancer patients using e.g. modern single-cell data, (ii) time-series analysis of multi-omics data from biomedical studies, (iii) (semi-)supervised analysis of extremely large-scale heterogeneous health data from Finnish biobanks, and (iv) novel cancer diagnostic methods using cell-free DNA. Your work would include familiarising yourself with one of these projects (based on your preference), contribute to developing statistical/deep learning methods, and apply them to exciting real-world data from our national or international collaborators. Applicants are expected to have good knowledge of machine learning/statistics, programming, and interest in developing/applying probabilistic methods for bioinformatics and biomedicine. Research work can be continued after the summer. For more information and relevant recent work, see http://research.cs.aalto.fi/csb/publications or contact Harri Lähdesmäki ([email protected]).

 

45) Field of study: Machine learning / computational biology

School / Department: School of Science, Department of Computer Science

Professor in charge of topic: Harri Lähdesmäki

Academic contact persons for further information on topic: Markus Heinonen ([email protected]), Mikko Arvas ([email protected]), Harri Lähdesmäki ([email protected])

Title of topic: Big Blood Data Prediction - machine learning for blood donation

Short task description:

Blood donation is a crucial life-saving voluntary activity, but recent research implies that it does not suite everybody. The goal of this project is to develop a classifier to identify persons who's health and wellbeing is minimally impacted by blood donation. A collaboration of Aalto University and Finnish Red Cross Blood Service (FRCBS) funded by European Blood Alliance aims to develop a prototype of such a classifier for the European blood banking community. Your job would be to initiate the development of the prototype under joint supervision of Aalto and FRCBS. In the project you will familiarize yourself with common blood biomarkers as well genomics data and modern machine learning techniques.  If successful the project could have immediate European wide impact and beneficially impact on health of numerous blood donors.

Research work can be continued after the summer.

The project requires good knowledge of mathematics, statistics, and programming (e.g. R, Python, Julia..) as well as interest in molecular biology and/or health.

For more information, see research group web page http://research.cs.aalto.fi/csb/ , https://www.bloodservice.fi/Research%20Projects

46) Field of study: Machine learning

School / Department: School of Science, Department of Computer Science

Professor in charge of topic: Harri Lähdesmäki

Academic contact person for further information on topic: Cagatay Yildiz ([email protected]), Markus Heinonen ([email protected]), Harri Lähdesmäki ([email protected])

Title of topic: Bayesian inference for nonparametric (deep) differential equations

Short task description:

Recently proposed nonparametric differential equation models make it possible to learn arbitrary continuous-time dynamics from data without any prior knowledge. These models can also be used to implement state-of-the-art deep learning methods. These high-capacity models can, however, suffer from over-fitting. Building on our research group's recent results, your work involves developing and implementing Bayesian inference methods (MCMC, variational inference) for robust inference of nonparametric differential equations. The work will include non-parametric probabilistic modelling, deep learning and automatic differentiation. The goal of the summer internship is to contribute to development of Bayesian methods for inferring non-parametric differential (deep) equation models from data and to implement these methods in e.g. TensorFlow.

The project requires good knowledge of machine learning, mathematics, statistics, and programming. Research work can be continued after the summer. For more information and relevant recent work, see http://research.cs.aalto.fi/csb/publications or contact any of the academic contact persons listed above.

47) Field of study: Machine learning

School / Department: School of Science, Department of Computer Science

Professor in charge of topic: Harri Lähdesmäki

Academic contact person for further information on topic: Markus Heinonen ([email protected]), Harri Lähdesmäki ([email protected])

Title of topic: Solving Fokker-Planck-Kolmogorov equations with deep learning

Short task description:

Recently in machine learning there has been emerging interesting towards dynamical models, such as RNN’s, ResNet’s or stochastic differential equation models. Fokker-Planck or Kolmogorov equations describe how the dynamics of such continuous-time stochastic dynamical systems evolve.

For general nonlinear dynamics the equations are intractable and warrant expensive numerical approximations. In this project we seek to sidestep the numerical approximations by solving the equation using deep machine learning instead, where we learn a neural network that predicts the system state. This work requires good knowledge of math and programming (Python preferred), as well as some knowledge of dynamical models. We will employ Tensorflow or other autodiff platforms for efficient exploration and learning of the neural networks.

Research work can be continued after the summer. For more information and relevant recent work, see http://research.cs.aalto.fi/csb/publications or contact any of the academic contact persons listed above.

48) Field of study: Machine learning

School / Department: School of Science, Department of Computer Science

Professor in charge of topic: Harri Lähdesmäki

Academic contact person for further information on topic: Cagatay Yildiz ([email protected]), Markus Heinonen ([email protected]), Harri Lähdesmäki ([email protected])

Title of topic: Auto-Differentiation Meets Differential Equations: A Tensorflow Extension

Short task description:

The automatic differentiation technique in machine learning has revolutionarised deep learning by allowing automatic gradients and optimisation of arbitrarily complex network structures. Recently there has been emerging interests in deep learning using inherently dynamical models such as ResNets or RNN’s, or even infinitesimal models such as continuous-time differential equations. However, the standard autodiff methods are inoptimal for these models. In this project we propose to extend automatic derivatives to differential equation system settings using existing forward/backward sensitivity analyses, and implement them into the Tensorflow package. This work requires good knowledge of math and Python programming, algorithmic mindset and knowledge of statistics or dynamic models being advantageous.

Research work can be continued after the summer. For more information and relevant recent work, see http://research.cs.aalto.fi/csb/publications or contact any of the academic contact persons listed above.

49) Field of study: Probabilistic modelling

School / Department: School of Science, Department of Computer Science

Professor in charge of topic: Aki Vehtari

Academic contact person for further information on topic: Aki Vehtari, [email protected]

Title of topic: Robust and Assisted Bayesian Modeling Workflows

Short task description:

You will participate in a research project in which we will develop theory and methods for assessing the quality of Monte Carlo and variational inference methods, and develop tools for a principled and robust Bayesian modeling workflow. To guarantee wide applicability of the project results in data science industry and academic research, the novel methods will be evaluated on a range of practical machine learning models and implemented as part of the leading open-source probabilistic programming systems. Prerequisite is knowledge Bayesian methods.

How to apply to AScI internship programme