Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Real-Time Deep Semantic SLAM
The goal of this project is to develop an object-level semantic SLAM system. The student will combine approaches from both SLAM and deep learning to develop a semantically-meaningful SLAM system.
Keywords: Semantic Segmentation; SLAM; Deep Learning; Autonomous Navigation; Computer Vision
The SLAM front-end will be done with a provided SLAM system. The student will have to focus on the backend of the SLAM system. Existing SLAM systems typically build a geometry-based map without semantic context. In this project, the student will focus on the strategy to fuse the geometry map with the semantic information extracted using state-of-the-art segmentation approaches to construct a semantically-meaningful and geometrically-accurate map. Along the way, the student will learn state-of-the-art approaches in both SLAM and deep learning.
The SLAM front-end will be done with a provided SLAM system. The student will have to focus on the backend of the SLAM system. Existing SLAM systems typically build a geometry-based map without semantic context. In this project, the student will focus on the strategy to fuse the geometry map with the semantic information extracted using state-of-the-art segmentation approaches to construct a semantically-meaningful and geometrically-accurate map. Along the way, the student will learn state-of-the-art approaches in both SLAM and deep learning.
- WP1: Research into state-of-the-art segmentation algorithms with low computational cost.
- WP2: Familiarisation with existing software for SLAM.
- WP3: Development and integration of your semantic segmentation pipeline with the SLAM output.
- WP4: Experimentation and evaluation of these algorithms against state-of-the-art approaches.
- WP5: Further optimization of the pipeline of WP3 to work in challenging scenarios.
- WP6: Final evaluation of the methods and report writing.
- WP1: Research into state-of-the-art segmentation algorithms with low computational cost. - WP2: Familiarisation with existing software for SLAM. - WP3: Development and integration of your semantic segmentation pipeline with the SLAM output. - WP4: Experimentation and evaluation of these algorithms against state-of-the-art approaches. - WP5: Further optimization of the pipeline of WP3 to work in challenging scenarios. - WP6: Final evaluation of the methods and report writing.
- Strong interest in computer vision.
- Experience in Linux is beneficial
- Strong analytical skills.
- Background in C++.
- Prior knowledge in SLAM or deep learning is preferred.
- Strong interest in computer vision. - Experience in Linux is beneficial - Strong analytical skills. - Background in C++. - Prior knowledge in SLAM or deep learning is preferred.
Your application must contain your most recent transcripts from bachelor and master studies. Please send the email to both supervisors. Dr. Zetao Chen - chenze@ethz.ch.
Your application must contain your most recent transcripts from bachelor and master studies. Please send the email to both supervisors. Dr. Zetao Chen - chenze@ethz.ch.