We just wrapped up the May 25, 2023 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 our monthly donation to charitable causes. The charity that received the highest number of votes this month was Wildlife AI! We were first introduced to Wildlife AI through the FiftyOne community. They are using FiftyOne to enable their users to easily analyze the camera data and create their own models. We loved their use case so much, we invited them to present, and as it turns out, they said yes and Victor Anton, Founder and CEO of Wildlife.ai, spoke at this very Meetup! We are sending this event’s charitable donation of $200 to Wildlife AI on behalf of the computer vision community.
Missed the Meetup? No problem. Here are playbacks and talk abstracts from the event.
Applying Computer Vision to Real Estate at Opendoor
There are many applications for computer vision in the real estate sector. Real estate is one of the few industries which still uses tools from the 2000s and can stand to benefit greatly from the advances in machine learning. In this talk, we will deep dive into Enricher – the computer vision pipelines that help Opendoor conduct virtual assessments of homes across the country at scale. Also discussed were some high level computer vision projects being worked on at Opendoor.
Shashwat Srivastava is a Senior Engineer on the Data Platform team at Opendoor. Opendoor is a real estate company which greatly simplifies the process of buying and selling your home. Over the 5 years, Shashwat worked on Data and Analytics at Opendoor on everything from data infrastructure to ETL pipelines to building backend services for real time serving of data. Before this role, he was a backend engineer at AppDynamics which he joined right after his undergrad at Carnegie Mellon.
Learn more about Opendoor’s computer vision work at the Opendoor Engineering and Data Science blog.
Q&A from the talk included:
- Do you use continuous training?
- Are there issues with the time delay between getting the video and the estimate on the cost of repairs?
- How do you monitor the results of your models?
- What was the rationale for doing a bytes conversion for Spark instead of other formats, say linearised or numpy?
- Is the home inspection done completely through computer vision or is having a human in loop still required?
- What about insulation and plumbing work? How do you inspect that?
YOLO-NAS – SOTA Object Detection Generated by NAS
Deci.ai is thrilled to announce the release of a new object detection model, YOLO-NAS. This model is a game-changer in the world of object detection, providing superior real-time object detection capabilities and production-ready performance. Eugene will discuss the new model, and the hardware-aware NAS approach.
Eugene Khvedchenia is a Deep Learning Engineer at Deci AI.
Q&A from the talk included:
- In knowledge distillation, how are the teacher network’s weights initialized and maintained?
- Is NAS applied only in the backbone? or in the decoder as well?
- Why is YOLO-NAS considered to be part of the YOLO family?
- YOLOv8 is licensed with limitations for commercial use. YOLO-NAS is Apache 2.0. What are the distribution limitations of YOLO-NAS?
- What kind of augmentations were tried in the training process?
- Did you consider pre-training with another dataset before training with COCO?
Wildlife Watcher: A Smart Wildlife Camera
Wildlife conservation is more critical now than ever, and monitoring biodiversity is key to protecting our planet. Wildlife.ai, an environmental non-profit, is using cutting-edge technology like artificial intelligence, computer vision and community collaboration to help biodiversity conservation. Their Wildlife Watchers program combines open-source, low-powered cameras with user-friendly software to make conservation accessible to everyone.
Victor Anton is Founder and CEO of Wildlife.ai, Victor is a wildlife biologist who works with research institutions, governments and communities around the world to translate data into improved conservation practices.
Q&A from the talk included:
- Depending on where you place your Wildlife Watcher, there would be more or less animal traffic through it. Given this, how do you deal with this for reproducibility?
- Has tinyML been useful for any of these projects?
- Can your camera be powered via a solar panel?
- Do you record and process sound, as well?
Join the Computer Vision Meetup!
Computer Vision Meetup membership has grown to over 4,000 members in just under a year! 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 computer vision and complementary technologies.
Join one of the 13 Meetup locations closest to your timezone.
- Ann Arbor
- New York
- San Francisco
- Silicon Valley
We have exciting speakers already signed up over the next few months! Become a member of the Computer Vision Meetup closest to you, then register for the Zoom.
Up next on June 8 at 10 AM Pacific we have the US and EMEA timezone-friendly Computer Vision Meetup happening with talks including:
- Redefining State-of-the-Art with YOLOv5 and YOLOv8 – Glenn Jocher (Ultralytics)
- Plug-and-Play Diffusion Features for Text-Driven Image-to-Image Translation – Narek Tumanyan & Michal Geyer (Weizmann Institute of Science)
- Re-annotating MS COCO, An Exploration of Pixel Tolerance – Jerome Pasquero & Eric Zimmermann (Sama)
Register for the Zoom here. You can find a complete schedule of upcoming Meetups on the Voxel51 Events page.
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 him over LinkedIn to discuss how to get you plugged in.
The Computer Vision Meetup network is 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.