View source for CISC849 F2023 Project

The final project is your opportunity to explore a mobile robot topic of your choosing.  You may use ROS or APIs/libraries/simulators (with appropriate citations for *any* code written by someone else), as long as you are adding some value -- i.e., not just downloading code and running it, or running somebody's training script on somebody's data.  Potential topics:

* Reinforcement learning or imitation learning in simulation (Isaac/Mujoco/Unity ML)
* Deep learning for visual object detection/segmentation
* Use AprilTags in the environment or attached to particular objects for mapping/localization/manipulation
* Some demonstration of lidar-based SLAM or localization on one of the robots
* Add a 3-D printed physical part to one of the robots to enable it to carry out a particular task
* Use the pan servo on <tt>yoshi</tt> to track and follow a person while using the lidar for obstacle avoidance
* Some other amazing idea that you have...

You may work alone or as part of a pair.  Previously-formed teams do NOT automatically continue for the project -- you and your partner need to affirm that you want to continue working together on the project.  So please send me an e-mail with your proposal and partner name, if applicable, as soon as possible but no later than Friday, November 17 so that I can give some feedback.  I don't just want a copy-paste from the list above but also at least a paragraph on specific libraries/software/datasets you might use and what deliverables you are targeting.  Time slots for 30-minute in-person demos (not in front of the class) on Monday, December 11 and Tuesday, December 12 (reading day) will be posted after Thanksgiving break.  Your code and write-up must also be submitted on the day of your demo.  Demos will take place in the lab.