Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Deep Learning for Semantic SLAM
The goal of this project is to train a deep neural network to segment robot sensor data (vision / LiDAR) into semantic scenes. The segmentation is then used to facilitate the SLAM process.
Keywords: Deep Learning, SLAM, Semantic Segmentation, Robotics
In this project we aim to leverage the challenge of Simultaneous Localization and Mapping (SLAM) to include semantics in the process. This shall help robots have a more human-like understanding of their environment, lead to efficient scene representations and improve the re-localization.
The main focus of this project is the development of a Deep Learning based semantic segmentation algorithm and integration with a state-of-the-art SLAM framework.
In this project we aim to leverage the challenge of Simultaneous Localization and Mapping (SLAM) to include semantics in the process. This shall help robots have a more human-like understanding of their environment, lead to efficient scene representations and improve the re-localization.
The main focus of this project is the development of a Deep Learning based semantic segmentation algorithm and integration with a state-of-the-art SLAM framework.
- Familiarize yourself with the current state of the art
- Implement state of the art solutions for semantic scene segmentation
- Train/evaluate the network on datasets of both synthetic and real sensor data
- Integrate semantic scene segmentation with SLAM system to facilitate re-localization
- Familiarize yourself with the current state of the art - Implement state of the art solutions for semantic scene segmentation - Train/evaluate the network on datasets of both synthetic and real sensor data - Integrate semantic scene segmentation with SLAM system to facilitate re-localization
- Good understanding of algorithmic challenges.
- Knowledge of C++, Python is mandatory.
- Knowledge of ROS is recommended.
- Knowledge in two of the following areas: SLAM, deep/machine learning and computer vision.
- Be curious about pushing the limits of today's robotics.
- Strong self-motivation and critical mind.
- Students from outside of D-MAVT (particularly, from D-INFK, D-ITET, D-PHYS, and D-MATH) are also highly encouraged to apply.
- Good understanding of algorithmic challenges. - Knowledge of C++, Python is mandatory. - Knowledge of ROS is recommended. - Knowledge in two of the following areas: SLAM, deep/machine learning and computer vision. - Be curious about pushing the limits of today's robotics. - Strong self-motivation and critical mind. - Students from outside of D-MAVT (particularly, from D-INFK, D-ITET, D-PHYS, and D-MATH) are also highly encouraged to apply.
If you are interested, please send your transcripts and CV to Abel Gawel (gawela@ethz.ch) and Anurag Sai Vempati (avempati@ethz.ch).
If you are interested, please send your transcripts and CV to Abel Gawel (gawela@ethz.ch) and Anurag Sai Vempati (avempati@ethz.ch).