The Voxel51 engineering team is thrilled to announce the general availability of FiftyOneOne 0.23.3 and FiftyOne Teams 1.5.4, which bring with them dozens of enhancements and fixes to streamline your computer vision workflows.
In a nutshell, here’s what’s new:
- Native Hugging Face and SuperGradients integrations
- Delete selected labels operator
- Monocular depth estimation tutorial
- API optimization
- Optimized cloud media export
This pair of releases include 30+ new features and fixes. Read on to learn more.
Wait, what’s FiftyOne?
FiftyOne is the open source machine learning toolset that 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.
Okay, but what’s FiftyOne Teams?
FiftyOne Teams extends FiftyOne with a GSuite-like experience for teams that want to collaborate on data stored in a centralized location with additional features like user permissions, dataset versioning, cloud-backed media, and enterprise security.
If this sounds interesting, read on! Then schedule a demo to learn more about FiftyOne Teams.
What’s new in FiftyOne 0.23.3
This release includes the following updates:
- Released a Hugging Face integration for running inference with
transformersmodels on your FiftyOne datasets!
- Released a SuperGradients integration for running inference with YOLO-NAS architectures!
- Primitive values in
DynamicEmbeddedDocumentlist fields are now displayed as comma-separated values (previously displayed as None) in the sample modal #3963
- Improved field visibility’s show metadata toggle #3926
- Fixed issues for unknown operator types and defaults #3851
- Miscellaneous saved view improvements #3974
- Fixed a bug where images in the sample modal errored when frame fields were added to video slices in mixed datasets #3966
- Fixed in-App sort by similarity for datasets with a color scheme #3966
- Fixed issues where media and other URLs were constructed incorrectly #3976
- Fixed keyboard navigation for dropdowns throughout the App #3965
- Added support for passing Hugging Face, Ultralytics, and SuperGradients models directly to brain methods #4004
- Added support to
register_run()for configuring whether run cleanup happens #3978
- Added support for passing model kwargs to
- Fixed issues with similarity searches on views and with pre-computed embeddings using the MongoDB backend
- Added dynamic batching to bulk writes like
- Added support for customizing progress bar rendering at method level #3979
- Include sample/frame singletons when clearing dataset cache via
- Fixed issues with embedded document field schemas #4002
- Added support for directly passing Ultralytics models to
- Added GPU support for OpenCLIP models #3986
- Added prompt embedding capabilities to OpenCLIP models #3960
- Added a builtin
ctx.selected_labelsformat to be consistent with other SDK methods #3998
What’s new in FiftyOne Teams 1.5.4
Includes all updates from FiftyOne 0.23.3, plus:
export()calls involving cloud-backed media
- Deployments with their
FIFTYONE_API_URIenvironment variable set will now display the API URI to users in the Teams App
- Improved debug logs by adding the head and tail of large results
motordependency to 3.3.0
- Fixed a regression when exporting cloud-backed media to CVAT for annotation
- Fixed an issue where API requests were not being prefixed with the correct proxy URL
- Fixed running
compute_similarity()over API connections with the MongoDB backend
Check out the release notes for a full rundown of additional enhancements and bugfixes in FiftyOne Teams 1.5.4.
Get involved in the FiftyOne open source community!
If you are working on computer vision use cases and unstructured data, the FiftyOne community is for you. There are tons of ways to get involved, for example:
FiftyOne Community Slack
With over 2,000 members, the community Slack channel is a great place to interact with the FiftyOne developers and exchange solutions with machine learning engineers doing computer vision in production.
To make it easy to catch the highlights, every Friday we recap interesting questions and answers from Slack in our Tips & Tricks blog series. Check out almost 300 tips and tricks in the archives like:
- 3D Detections – FiftyOne Tips and Tricks
- Exploring Polylines – FiftyOne Tips and Tricks
- Creating Pose Skeletons from Scratch – FiftyOne Tips and Tricks
- Dynamic Groups – FiftyOne Tips and Tricks
- Understanding Grouped Datasets
Computer Vision and AI, Machine Learning, and Data Science Meetups
Voxel51 sponsors 13 virtual Computer Vision Meetups and 12 AI, Machine Learning and Data Science Meetups around the world with almost 18,000 members. (To join, visit the Meetup links and scroll down to find the location friendliest to your time zone.)
The 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. Our next Meetup is happening this Thursday:
- Lightning Talk: Next-Generation Image/Video editor Built with Generative AI – Mihail Eric at Storia AI
- SANPO: A Scene Understanding, Accessibility, Navigation, Pathfinding, Obstacle Avoidance Dataset – Kimberly Wilber at Google Research
- Setmlvis: Object Detection Comparison Using Set Visualization – Liudas Panavas at Northeastern University
- How well does the Segment Anything Model work on a Fisheye lens? – Nahid Alam at Cisco Meraki
FiftyOne on GitHub
If you want to start contributing to the FiftyOne project resolving issues, reporting bugs or making enhancements to the Docs, check out these resources:
Industry Spotlight: Computer Vision in Sports
Check out the fourth installment of Voxel51’s computer vision industry spotlight blog series. In this series, we highlight how different industries — from construction to climate tech, from retail to robotics, and more — are using computer vision, machine learning, and artificial intelligence to drive innovation. We’ll dive deep into the main computer vision tasks being put to use, current and future challenges, and companies at the forefront.
In this edition, we’ll focus on sports! Read on to learn about computer vision in the sports industry. Check out the “How Computer Vision Is Changing Sports” to get the details!