What we’re about
Meet and discuss applications and algorithms related to machine learning and AI. Use our speaker sign-up form.
Other Cool Melbourne Meetups
MLOps: https://www.meetup.com/melbourne-mlops-community1/
BeerOps: https://www.linkedin.com/company/beerops/?originalSubdomain=au
Melbourne Automation Meetup: https://www.meetup.com/melbourne-automation-meetup/
Upcoming events
5
Seeing+Doing: Vision Language Action policies, SO-101, & Collective Computation
912 Collins St, Docklands, VI, AUThe MLAI Meetup is a community for AI researchers and professionals which hosts monthly talks on exciting research. Our format is:
- 6:00 - 6:20: Socializing
- 6:20 - 6:40: Announcements and AI news
- 6:40 - 7:40: Talk(s) and Q&A
- 7:40 - 8:00 Networking
- 8:00: Head to the nearest pub for dinner
We have a great event this month, with a long talk from Andy Gelme and a short talk from Ahmet Sekercioglu!
Andy Gelme: "Seeing + Doing: Using Vision Language Action (VLA) policies with LeRobot SO-101 arms"
Abstract: Robotics is now turning perception into action using foundation models.
Vision Language Action (VLA) policies typically incorporate transformer
architectures by encoding multimodal embeddings, such as vision and text instructions, whilst tokenising robot state and action sequences, e.g
motor joint trajectories (tokens) during training ... and decoding
output actions (tokens) mapped to motor joint sequences.This talk introduces Hugging Face’s open-source software and hardware
LeRobot project and the affordable SO-101 robotic arms. Key concepts
will be explained ... and there will be code !The entire workflow will be covered ... from data capture, uploading and
sharing datasets, training VLA policies and running those learned policies.There will be a live hardware demonstration of the LeRobot SO-101 arms.
What could possibly go wrong ?Speaker Bio: Andy is an A.I Systems Engineer and Hacker At Large: https://github.com/geekscape
Ahmet Sekercioglu: "Collective Computation in Neuronlike Circuits"
Abstract: When I learned that Hopfield and Hinton won the Nobel Prize for their work on artificial neural networks, I remembered Tank and Hopfield's 1987 Scientific American article "Collective Computation in Neuronlike Circuits." While searching for a PhD topic in 1990, I came across this article and found it fascinating. Unfortunately, I liked the Fuzzy Logic concept more and chose to focus my PhD thesis on that field instead. In hindsight, perhaps I should have chosen artificial neural networks!
I recently revisited the Scientific American article. It is beautifully written and I think it is still relevant today. It is particularly valuable for anyone wanting to understand the foundational concepts that eventually led to technologies like ChatGPT.
In this talk, I will walk through the article, explain the key concepts that Tank and Hopfield present, and walk through one example to show how these early ideas connect to modern AI.
Speaker Bio: I am now semi-retired. I started my career as a research engineer in the early 1980s, worked on various projects and in the early 1990s switched to academia. I finished full-time work in 2019.
For details: https://sekerci.info/~ahmet
48 attendees
Past events
135
