We just wrapped up the August ‘24 AI, Machine Learning and Computer Vision Meetup, and if you missed it or want to revisit it, here’s a recap! In this blog post you’ll find the playback recordings, highlights from the presentations and Q&A, as well as the upcoming Meetup schedule so that you can join us at a future event.
First, Thanks for Voting for Your Favorite Charity!
In lieu of swag, we gave Meetup attendees the opportunity to help guide a $200 donation to charitable causes. The charity that received the highest number of votes this month was Heart to Heart International, an organization that ensures quality care is provided equitably in medically under-resourced communities and in disaster situations. We are sending this event’s charitable donation of $200 to Heart to Heart International on behalf of the Meetup members!
Missed the Meetup? No problem. Here are playbacks and talk abstracts from the event.
GenAI for Video: Diffusion-Based Editing and Generation
Recently, diffusion-based generative AI models have gained popularity due to their wide applications in the image domain. Additionally, there is growing attention to the video domain because of its ubiquitous presence in real-world applications. In this talk, we will discuss the future of GenAI in the video domain, highlighting recent advancements and exploring its potential and impact on video editing and generation. We will also examine the challenges and opportunities these technologies present, offering insights into how they can revolutionize the video industry.
Speaker: Ozgur Kara is a PhD student in the Computer Science Department at the University of Illinois at Urbana-Champaign. He earned his Bachelor’s degree in Electrical and Electronics Engineering from Boğaziçi University. His research focuses on generative AI and computer vision, particularly on generative AI and its applications in video.
Resource Links
- Project: RAVE: Randomized Noise Shuffling for Fast and Consistent Video Editing with Diffusion Models
- Paper
- Code
Q&A
- For sampling video data for training, how do you choose which frames to sample?
- When you shuffle the frames, does this put a limit on the number of frames you can process to get it processed quickly?
- Are you able to speak to any research going on for 3D object generation for AR/VR use cases?
Evaluating RAG Models for LLMs: Key Metrics and Frameworks
Evaluating the model performance is the key for ensuring effectiveness and reliability of LLM models. In this talk, we will look into the intricate world of RAG evaluation metrics and frameworks, exploring the various approaches to assessing model performance. We will discuss key metrics such as relevance, diversity, coherence, and truthfulness and examine various evaluation frameworks, ranging from traditional benchmarks to domain-specific assessments, highlighting their strengths, limitations, and potential implications for real-world applications.
Speaker: Abi Aryan is the founder of Abide AI and a machine learning engineer with over eight years of experience in the ML industry building and deploying machine learning models in production for recommender systems, computer vision, and natural language processing—within a wide range of industries such as ecommerce, insurance, and media and entertainment. Previously, she was a visiting research scholar at the Cognitive Sciences Lab at UCLA where she worked on developing intelligent agents. Also, she has authored research papers on AutoML, multi agent systems, and LLM cost modeling and evaluations and is currently authoring LLMOps: Managing Large Language Models in Production for O’Reilly Publications.
Resource links
Q&A
- What are the best strategies for integrating and performing evaluations of Large Language Models (LLMs) within a CI/CD pipeline, specifically in the context of Retrieval-Augmented Generation (RAG) workloads?
- How do RAG systems prioritize the information retrieved and incorporated into th prompt’s context, or do they? And how does the LLM recognize that priority?
- Additionally, how can we define and implement key metrics to determine whether a model should be deployed or not, ensuring it meets the necessary performance, reliability, and safety standards for RAG and information retrieval applications?
- How is the Galileo RAG? Any information on that?
- What if the RAG system retrieves conflicting data?
Why You Should Evaluate Your End-to-End LLM applications with In-House Data
This task discusses end-to-end NLP evaluations, focusing on key areas, common pitfalls, and the workings of production evaluation systems. It also explores how to fine-tune in-house LLMs as judges using custom data for more accurate performance assessments.
Speaker: Mahesh Deshwal is a Data Scientist and AI researcher with over 5.5 years of experience in using ML and AI to solve business problems, particularly in Computer Vision, NLP, recommendation, and personalization. As the author of the paper PHUDGE and an active open source contributor, he excels in delivering end-to-end solutions, from user requirements to deploying scalable models using MLOps.
Resource links
Join the AI, Machine Learning and Computer Vision Meetup!
The goal of the Meetups is to bring together communities of data scientists, machine learning engineers, and open source enthusiasts who want to share and expand their knowledge of AI and complementary technologies.
Join one of the 12 Meetup locations closest to your timezone.
- Athens
- Austin
- Bangalore
- Boston
- Chicago
- London
- New York
- Peninsula
- San Francisco
- Seattle
- Silicon Valley
- Toronto
What’s Next?
Up next on Aug 15, 2024 at 2 PM BST / 6:30 PM IST , we have three great speakers lined up!
- SyntStereo2Real: Edge-Aware GAN for Remote Sensing Image Translation while Maintaining Stereo Constraints – Vasudha Venkatesan, University of Freiburg/ex-German Aerospace Centers
- Elevating Security with Data-Centric AI: A Comprehensive Approach to Surveillance- Daniel Gural, ML Engineer at Voxel51
- How to Develop Data Science Projects with Open Source – Jay Cui – MLOps Engineer
Register for the Zoom here. You can find a complete schedule of upcoming Meetups on the Voxel51 Events page.
Get Involved!
There are a lot of ways to get involved in the Computer Vision Meetups. Reach out if you identify with any of these:
- You’d like to speak at an upcoming Meetup
- You have a physical meeting space in one of the Meetup locations and would like to make it available for a Meetup
- You’d like to co-organize a Meetup
- You’d like to co-sponsor a Meetup
Reach out to Meetup co-organizer Jimmy Guerrero on Meetup.com or ping me over LinkedIn to discuss how to get you plugged in.
—
These Meetups are sponsored by Voxel51, the company behind the open source FiftyOne computer vision toolset. FiftyOne enables data science teams to improve the performance of their computer vision models by helping them curate high quality datasets, evaluate models, find mistakes, visualize embeddings, and get to production faster. It’s easy to get started, in just a few minutes.
