I was going through the JdeRobot GSoC website and the idea of developing Computer Vision exercises for JdeRobot Academy sounded really interesting to me. I have done courses in Computer Vision and have plenty of experience working with OpenCV. You can find a compilation of assignments from my graduate CV course here. I think this would be a great opportunity for me to learn ROS2, since I've only ever worked with the first version. Kindly let me know a bit more about what you have in mind (whether the assignments would tilt more towards traditional computer vision techniques, or would the focus be more on CNNs/Deep Learning).
I realize I'm a bit late in starting up. But I'm still confident that I can get a good proposal ready by the end of this week. Let me know if you need any more information from my side.