About topic

• Participate in Robotour competition

• Develop a real-time road segmentation algorithm that processes RGB camera frames onboard a ground robot

• Fine-tune a pre-trained model on a custom dataset and deploy it within the robot’s perception pipeline

• Optionally fuse image-based segmentation outputs with 3D LiDAR or multispectral sensors to improve robustness in varied environments

• Part of a larger Autonomous Navigation system within the Robotour ground-robot competition

• You’ll work in a multidisciplinary team to integrate perception, planning, and control

Suitable for: Anyone passionate about Computer Vision, AI, and Mathematics, self-motivated learners with analytical thinking and a collaborative mindset

Requirements: Basic programming skills in Python, interest in research, prior experience with image classification or segmentation is an advantage

What courses may be useful? Machine Learning; Computer Vision; Autonomous Navigation.

Work can be adapted towards: bachelor or master thesis, internship project, or standalone software

Are you interested?

Contact hulchvse@student.cvut.cz!