Real-time image recognition using machine learning without dataset

About the job
This project is carried out at ÅF with a supervisor from ÅF and in collaboration with KTH Formula Student. The project aims to support two types of cars, ÅF has a small 1:12 scaled autonomous car for rapid prototyping of functionality and KTH formula student are designing an autonomous, self-driving electric formula car to participate in the largest engineering student competition in the world, Formula Student. http://www.kthformulastudent.se/
Description
Most machine learning algorithm are trained using a labeled dataset where the samples in the dataset has already been classified.The idea is to be able to train a learning algorithm such as a neural network without using a dataset or labeled data. Instead the algorithm is trained using significant features like the shape of the objects that we want to detect on the road. The importance of good real time performance is emphasised and results should be evaluated towards results with other algorithms. The first type of object/shape to be trained/recognized should be traffic cones for the autonomous car competition.

Who are you?
The thesis is at master's level, scheduled to be started by Q1-2019, suitable for two persons and will be performed mainly at ÅF's office in Solna. We do also accept single candidates and can provide candidate matching.
This degree project will require good software knowledge and skills in programming. We believe you have knowledge in:

C++ and/or Python
Machine learning and or image recognition

A requirement is that you have good knowledge in English, both oral and written. Knowledge in Swedish is meritorious.

We offer
ÅF is Swedens largest technical consultancy. When doing your master thesis at ÅF you will be a part of our business, participate in internal activities, get to know other ÅF consultants and see what we do for our customers.

About the company
ÅF is an engineering and design company within the fields of energy, industry and infrastructure. We create sustainable solutions for the next generation through talented people and technology. We are based in Europe and our business and clients are found all over the world.

ÅF Making Future.