In-class Presentation Guidelines
In-class Presentation Guidelines
The main goal of the in-class presentation is for you to explore in-depth a research topic/paper in robotic manipulation of your own interest, learn how to read and analyze a paper, and perform an academic presentation.
You can propose your own or choose from this reading list I prepared. This reading list includes some interesting papers on various topics in robotic manipulation. Submit a list of your top 3 choices (topics or papers) before the deadline, preferably as soon as possible, as this will give us more time to find the most interesting material for you and resolve any potential conflict.
The order of presentations will be determined based on the topics (similar topics will be grouped).
Please read through the guidelines below.
There are three options:
You can present one paper in depth.
You can learn and present some classic references/techniques in robotic manipulation (e.g., “I want to understand how to model contacts”, “I want to learn how impedance control works”, or “I want to learn about reinforcement learning techniques for in-hand manipulation”).
You can give an overview of the research in a group/individual/small area (e.g., “what are people working on in the Foam Robotics Lab at CMU”, “an overview of Matt Mason’s research”, or “review sim-to-real gap in manipulation or review planning and learning for multi-object manipulation”).
You are welcome to select a paper or get inspiration from the reading list. In general, you are free to choose any topic related to robotic manipulation, here are some examples:
Planning or(and) learning in robotic grasping
Planning for dexterous manipulation
Imitation learning/learning from demonstration/reinforcement learning in manipulation
LLMs/VLMs and manipulation
Survey of recent hand designs
Benchmarks in manipulation
Multi-object manipulation
Tool-use for manipulation
Soft robotic hands / Tendon-drive hands / Prosthetic hands
Compliance
Contact modeling
Teleoperation techniques for robotic manipulation
the use of VR in robotic manipulation
Evolution of human hands / Manipulation skill development in young children
Non-prehensile manipulation behaviors in human manipulation
Direct-drive techniques in robotic manipulation
Feel free to select a topic from this list or be creative and use this list as inspiration to explore a subject that truly interests you! You are also welcome to present a paper not directly focused on manipulation. In this case, your presentation should highlight the paper's connection and potential applications for manipulation.
If you don’t already have a topic or paper in mind, a great starting point is to check the reading list or browse recent journals and conferences. For journals, consider looking at IJRR, TRO, Science Robotics, IEEE-RAL, etc. For conferences, you might explore RSS, ICRA, IROS, CoRL, ISRR, WAFR, Humanoids, SIGGRAPH, etc. If you have some preliminary ideas but need help narrowing down where to look, feel free to reach out for guidance. If you have a topic in mind but aren’t sure which paper(s) to focus on, we can work together to choose suitable reading materials.
You are required to give one presentation during the semester. The presentation should typically last 25 to 30 minutes, followed by a 5-10 min discussion. Longer presentation time is allowed upon request.
Aim for about one minute per slide. Keep text minimal and easy to read, and use images, plots, videos, or diagrams wherever they enhance your explanation. Please share your presentation material with me either before or after your presentation.
If you are presenting a specific paper, here are key points to cover. You don’t have to follow this exact order. For a collection of papers or a survey, your approach will differ slightly.
Introducing the authors (briefly): Who are the authors? What do we know about their lab? What else have they done? Mostly, we could focus on the first author(s) and the last author (usually the advisor).
Introducing the paper: What is the goal of the paper, and why is it interesting and important? This is often a good place to show a preview of results to give us an idea of what’s coming.
Related work: What has already been done in this area? What background are the authors building on?
Specific contribution: What gap is this paper addressing, and what is the authors’ stated contribution?
Techniques: How did the authors solve the problem? Highlight anything particularly interesting or insightful. If some parts were hard to understand, take this opportunity to clarify them—use the board or additional materials to explain.
Experiments: How did the authors design the experiments? Are the experiments thorough enough to verify their assumptions or demonstrate that they solve the problem?
Results: Present the results and help the audience focus on what is most interesting and important, especially if the paper contains many plots or data.
Future Work/Extension: If the paper is not recent, what follow-up work has it inspired or extended, and can this idea still inspire our current research? If the paper is recent, what potential future directions could it lead to?
Discussion: Share your thoughts about the paper. Evaluate the paper in validity and significance. Evaluate the quality of the work (methodology, techniques, accuracy, assumptions). Here are some questions to think about:
Are the authors addressing an interesting problem?
How well did they solve it?
Are the results useful, and for whom?
What might you improve or do differently?
What important problems remain unsolved, and where do you see the field heading?
Feel free to go beyond these questions—use them as a starting point to organize your thoughts. You are encouraged to ask the audience probing and bold questions to spark the discussion!
A good article on how to read a research paper and how to efficiently survey a new field:
I highly recommend you check out these video tutorials for academic presentations: