Context
Collecting indoor trajectories for an Indoor Positioning System (IPS) can be challenging due to the various challenges that can affect the accuracy and reliability of the data. Some of those problems are related to the heterogeneity of IPS, which use a different types of sensors and technologies. Another challenge is the privacy concerns surrounding the collection of indoor trajectories, which can limit the availability of real-world data for training and evaluation purposes. These challenges can make it difficult to accurately and reliably collect indoor trajectories.
Objective
The goal of this project is to research, design and implement some generative models for generating synthetic indoor trajectories. For this project, the student will train the models on a dataset of real-world indoor trajectories to generate synthetic trajectories that closely mimic the patterns and characteristics of the real-world data.
Requirements
- Solid programming skills, particularly in Python or Java,
- Good problem-solving skills and ability to work in a team environment,
- Ability to work independently and manage time effectively,
- Good written and oral communication skills,
- Some experience with Machine Learning libraries (TensorFlow, Scikit-learn, Pytorch, etc…) and algorithms is a plus.