🔧 Recapping the AI, Machine Learning and Computer Meetup — October 10, 2024
Nachrichtenbereich: 🔧 Programmierung
🔗 Quelle: dev.to
We just wrapped up the October ‘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.
How Renault Leveraged Machine Learning to Scale Electric Vehicle Sales
In 2019, Renault sought a scalable solution to estimate the cost of home charging station installations for electric vehicle buyers. A machine learning solution using satellite images and a shortest-path algorithm was developed to automate this process. Despite challenges, the optimized solution was deployed as a cloud-based API, enabling Renault to scale their EV sales from 50,000 in 2019 to over 220,000 in 2022.
Speaker: With a PhD in Physics, Vincent Vandenbussche has over a decade of experience deploying scalable machine learning solutions for leading companies like Renault and Chanel. He is also passionate about sharing his expertise through Medium posts and his book, The Regularization Cookbook.
Q&A
- Are you using imagery from Maxar, Sentinel, Airbus, and/or another source?
- What is meant by image binning?
- Did you explore semantic planning to address the Djikstra issue?
Resource Links
- The Regularization Cookbook
- Vincent’s blogs on Medium
RGB-X Model Development: Exploring Four Channel ML Workflows
Machine Learning is rapidly becoming multimodal. With many models in Computer Vision expanding to areas like vision and 3D, one area that has also quietly been advancing rapidly is RGB-X data, such as infrared, depth, or normals. In this talk we will cover some of the leading models in this exploding field of Visual AI and show some best practices on how to work with these complex data formats!
Speaker: Daniel Gural is a seasoned Machine Learning Evangelist with a strong passion for empowering Data Scientists and ML Engineers to unlock the full potential of their data. Currently serving as a valuable member of Voxel51, he takes a leading role in efforts to bridge the gap between practitioners and the necessary tools, enabling them to achieve exceptional outcomes. Daniel’s extensive experience in teaching and developing within the ML field has fueled his commitment to democratizing high-quality AI workflows for a wider audience.
Q&A
- If a 4th channel is lidar, it’s more sparse than the RGB resolution usually. Have you seen this in use?
- Are there data or modeling considerations you can mention about handling this "sparse 4th channel" case?
- When building the model, while feeding in more information, is there a general concern for overfitting the model and straying away from generalization? With imaging, I used to believe that more data is better to train the model, so in this case, would more information refer to more data in this context, please?
- Would it be correct to see the X in RGB-X model as different channels in a CNN model?
Resource Links
- Tutorial: Monocular Depth Estimation with FiftyOne
- Paper: Sapiens: Foundation for Human Vision Models
Elasticsearch is for the Birds: Identifying Feathered Friend Embedding Images as Vector Similarity Search
Search no longer has to be a traditional term / frequency-inverse document frequency slog. The current trend of machine learning and models has opened another dimension for search, quite literally. In this talk we’ll cover “classic” search and its inherent limitations, what a model is and how you can use it.
Next, we’ll look at how to perform vector search or hybrid search using images of birds in Elasticsearch, generate embeddings from images and then use the techniques we learned to propose the most probable similar images after uploading our own image. We’ll close the talk with an overview of various enhancements to increase the performance and usability of your searches.
Speaker: Justin Castilla is a Senior Developer Advocate at Elastic based in Seattle. His main focus is education and developer empowerment, and enjoys sharing knowledge and learning experiences with everyone.
Resource Links
- Demo Repo on GitHub
- Slides from the talk
Using Elasticsearch Vector Search in FiftyOne
The short demo focusses on how to leverage Elastic’s vector search search capabilities for computer vision use cases using the FiftyOne open source library.
Speaker: Steve Pousty is a Developer Advocate at Voxel51.
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 Oct 24, 2024 at 2:00 PM BST / 6:30 PM IST, we have two great speakers lined up!
Accelerating Machine Learning Research and Development for Autonomy- Guillaume Rochette, Oxa
Pixels Are All You Need: Utilizing 2D Image Representations in Applied Robotics- Brent Griffin, Voxel51
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.
...