CEVT is an innovation and development company at the forefront of mobility. Modular development, ground-breaking virtual engineering, software development and continuous innovation enable us to deliver world-class technology. Today we deliver to different automotive brands in the Geely family such as Geely Auto, Volvo Cars, and Lynk & Co and are based at Lindholmen in Gothenburg.
Master Thesis Project
Title: Scene Change Detection
Subject: Machine learning, Algorithm development and analysis
Scene change detection (SCD) refers to the task of localizing changes and identifying change-categories given two scenes. A scene can be either an RGB (+D) image or a 3D reconstruction (point cloud). If the scene is an image, SCD is a form of pixel-level prediction because each pixel in the image is classified according to a category. On the other hand, if the scene is point cloud, SCD is a form of point-level prediction because each point in the cloud is classified according to a category.
Some example benchmarks for this task are VL-CMU-CD, PCD, and CD2014. Recently, more complicated benchmarks such as ChangeSim, HDMap, and Mallscape got released. Models are usually evaluated with the Mean Intersection-Over-Union (Mean IoU), Pixel Accuracy, and F1 metrics.
CEVT’s ML team has been exploring SCD (scene change detection) algorithms with the aim of localizing changes in the image plane given two frames (t0, t1) taken at separate times.
The overall aim of the thesis work is to explore the possibilities of learning the changes with a single image and a pure semantic segmentation approach or even with a contrastive learning approach and compare it to our in-house developed approach.
- DL segmentation literature review and analysis
- Development steps:
- Data inspection
- Data loading
- Data pre-process
- Training relevant architectures
- Benchmark different approaches and recommend the best scene object detection methodology based on that
We are searching for passionate and team-oriented students with the following background:
- Master’s student in Computer Science, Data Science, Electrical Engineering, Engineering Physics or similar.
- Passion in machine learning
- Proficient skills in Python and DL frameworks (Pytorch or Tensorflow)
- Some knowledge in computer vision deep learning architectures
- Familiar with Linux environments
Number of students: 2, please apply with CV and cover letter.
Starting date: January 2022. The duration of the thesis work is 20 weeks.
For more information please contact:
Dorian Valverde Baspineiro, Product Owner Cloud, +46 729 888 029, firstname.lastname@example.org
Mikel Broström, AI/ML Engineer Cloud, +46 738 992 210, email@example.com
Last application date: 2021-10-29
Apply today. We will perform ongoing selection during the application period. We look forward to hearing from you!
Please note that due to GDPR regulations we can only accept applications sent through the recruitment system, not via email or other channels.