Welcome to the monthly blog series where we bring you up to speed on recent happenings in the FiftyOne community and celebrate noteworthy milestones. 🙌 🚀
Community Spotlights
We love hearing how FiftyOne helps you solve challenges and reach new heights! Curious what sorts of use cases are possible with FiftyOne? Here’s a highlight from the open source FiftyOne community.
Protex AI’s AI-powered technology enables businesses to gain greater visibility of unsafe behaviors in their facilities. The privacy-preserving platform plugs into existing CCTV infrastructure and uses its computer vision technologies to capture unsafe events autonomously in settings such as warehouses, manufacturing facilities, and ports.
“At Protex AI we develop a platform to monitor worker health and safety using existing CCTV infrastructure. Our core computer vision technologies are object detection, classification, and pose estimation. We use FiftyOne as a vital component in our pipeline to validate dataset annotations, intelligently subsample datasets to ensure balance, and also to visualize and debug model predictions to assess accuracy.”
— Patrick Rowsome — Lead Computer Vision Engineer, Protex AI
Community Rewards
Is your organization using FiftyOne to solve interesting computer vision problems? Share your success story and claim a box of community rewards as a thank you!
Product Releases
In September there were a few point releases to both the open source and Teams versions of FiftyOne. Here are the links to the releases learn more:
Open Source FiftyOne
- FiftyOne 0.22.0 – Sept 20
FiftyOne Teams
- FiftyOne Teams 1.4.1 – Sept 21
- FiftyOne Teams 1.4.0 – Sept 20
What’s new?
September’s releases brought us three big exciting features!
Dataset versioning
Dataset versioning in FiftyOne Teams allows you to track the lineage of your data, view the exact version of a dataset on which a given model was trained, protect against accidental data modification, and much more. With dataset versioning, you can trust FiftyOne Teams as the single source of truth for your organization’s data.
Native Ultralytics integration
FiftyOne now includes a native Ultralytics integration, making it easy to train the latest Ultralytics models like YOLOv8 on your FiftyOne datasets with just a few simple lines of code.
Delegated operations
Last month’s FiftyOne Teams also added a powerful new Delegated Operations feature to FiftyOne’s Plugin framework that allows you to schedule builtin and/or custom tasks from within the App that are executed on a connected workflow orchestrator like Apache Airflow.
Why is this awesome? Your AI stack needs a flexible data-centric component that enables you to organize and compute on your data. With Delegated Operations, FiftyOne becomes both a dataset management/visualization tool and a workflow automation tool that defines how your data-centric workflows like ingestion, curation, and evaluation are performed. In short, think of FiftyOne Teams as the single source of truth on which you co-develop your data and models together.
FiftyOne on GitHub
GitHub is home to the open source FiftyOne project. Here’s the latest snapshot of what’s happening in the FiftyOne GitHub repo:
- Total stars: 5,000+
- Total contributors: 74
- Total used by: 395
- Total forks: 424
- Total issues closed so far: 886
FiftyOne Community Slack
The FiftyOne Community Slack channel is where you can join more than 2,100 machine learning engineers and data scientists using FiftyOne to improve the quality of their computer vision data and build better models. Ask questions, answer questions, or simply follow along with the discussion!
To make it easy to catch the highlights, every Friday we recap interesting questions and answers from Slack in Tips & Tricks blog series. Recent posts include:
- 3D Detections – FiftyOne Tips and Tricks – September 29th, 2023
- Exploring Polylines – FiftyOne Tips and Tricks – September 22nd, 2023
- Creating Pose Skeletons from Scratch – FiftyOne Tips and Tricks – September 15, 2023
- Dynamic Groups – FiftyOne Tips and Tricks – Sep 8, 2023
- Understanding Grouped Datasets – FiftyOne Tips and Tricks – Sep 1, 2023
Computer Vision Meetups
Voxel51 sponsors 13 virtual Computer Vision Meetups and 12 AI, Machine Learning and Data Science Meetups around the world. (To join, visit the Meetup links and scroll down to find the location friendliest to your time zone.)
The Computer Vision Meetups are geared towards data scientists, machine learning engineers, and open source enthusiasts who want to expand their knowledge of computer vision and complementary technologies. We put an emphasis on open source software, and speakers who are computer vision practitioners or academics doing research in the field. This month’s Meetups include:
Oct 12 Computer Vision Meetup
- Bridging the Gap: Advancing Civil Engineering Inspections with Computer Vision – Johannes Flotzinger, Civil Engineer & Research assistant at Universität der Bundeswehr München
- Adapting to Change: Foundation Models, APIs, and the Past, Present and Future of AI Development – Pietro Bolcato at Kittl
Recapping Previous Meetups
If you missed any of last month’s Meetups, you can get the executive summaries and links to the video playbacks here:
- Recapping the Computer Vision Meetup — Sept 14
- Recapping the AI, Machine Learning, and Data Science Meetup — Sept 7, 2023
Upcoming Computer Vision Events
In addition to Meetups, we invite you to join us for one or more of these upcoming events:
- Oct 12 – October Virtual Computer Vision Meetup
- PyTorch Conference 2023
- Oct 25 – Getting Started with FiftyOne Workshop
- Nov 15 – FiftyOne Plugins Workshop: Authoring Data-centric AI Applications
New Docs, Blogs, Videos, and Tutorials
We want everyone to be successful with FiftyOne, and one of the ways we try to do that is by publishing resources that you might find helpful and handy. Here’s a list of some of the new documentation, blogs, videos, tutorials, integrations, and cheat sheets that you may want to check out.
Blogs
- Badger: Custom GitHub Badges Made Easy
- 3 Reasons to Visit Voxel51 at ICCV23!
- Zero-shot Prediction Plugin for FiftyOne
- Mastering Drone Data Training
- Optical Character Recognition with PyTesseract
- Navigating the Road Ahead
- CaFFe: Calving Fronts and Where to Find Them
- Double Trouble: Eliminate Image Duplicates with FiftyOne
- FACET: A Benchmark Dataset for Fairness in Computer Vision
Videos
- Filtering computer vision data with FiftyOne
- FiftyOne Computer Vision Tips and Tricks: Heatmaps
- FiftyOne Computer Vision Tips and Tricks: 3D Detections
- What is the FiftyOne computer vision toolset?
- FiftyOne Tips and Tricks: Creating Pose Skeletons from Scratch
- Computer Vision Meetup: Visualizing Defects in Amazon’s ARMBench Dataset
- Computer Vision Meetup: Optimizing Dist Workloads for Stable Diffusion
- Computer Vision Meetup: From Model to the Edge, Putting Your Model into Production
- Computer Vision Meetup: ARMBench: An Object-Centric Benchmark Dataset
- FiftyOne Computer Vision Tips and Tricks: Dynamic Groups
- Computer Vision Meetup: EgoSchema: A Dataset for Truly Long-Form Video Understanding
- Computer Vision Meetup: Monitoring Large Language Models (LLMs) in Production
- Computer Vision Meetup: Neural Residual Radiance Fields
- FiftyOne Teams: Dataset Versioning
- FiftyOne Computer Vision Tips and Tricks: Grouped Datasets
Voxel51’s Commitment to Open Source and Community
Open source, transparency, and giving back to the computer vision community is what we are all about! Whether it’s developing the open source FiftyOne computer vision toolset to help engineers and data scientists build high-quality datasets and models, sponsoring Meetups to help members boost their computer vision knowledge, or giving to charitable causes on behalf of the community, Voxel51 is committed to bringing transparency and clarity to the world’s data.
