Roboflow Changelog: March 2022

Spring is finally here and there is computer vision in the air. Each month, we share a recap of product and company updates; if you missed last month's edition of the Roboflow Changelog, you can find it here.

A note on licensing: YOLOv5 is released under AGPL-3.0, which affects commercial use:

• On Roboflow cloud: a commercial license is included for every user, including the free plan.
• Self-hosted: you need an Enterprise license add-on, or you must open-source your application under AGPL-3.0.
• Permissive alternative: RF-DETR (opens in a new tab) is Apache 2.0, free for commercial use anywhere.

Roboflow Organize and Annotate

• Launched Polygonal Annotation tool
• Launched Multi-label Classification import and labeling
• UX enhancements when creating a workspace
• Mobile responsiveness updates
• Support binary file uploads in the upload API
• Added user moderation tool
• Added support for importing Sagemaker Ground Truth annotation files on all plans
• Added "filter by class" view for Classification projects (click the class name from Health Check)

Roboflow Train and Deploy

• Launched Multi-label Classification Training
• Launched Multi-label Classification Hosted Inference API
• Inference APIs now support multipart requests
• Updated webcam demo with per-class confidence thresholds, camera flip option, and custom code integration
• Support for GPU inference server offline mode when network connection is completely disconnected

Roboflow Universe (opens in a new tab)

• Added 5 Popular Industry pages (Self Driving (opens in a new tab), Manufacturing (opens in a new tab), Gaming (opens in a new tab), Agriculture (opens in a new tab), Sports (opens in a new tab))
• Inference test widget now supports Single-Class and Multi-Class Classification projects
• SEO page improvements
• Mobile responsiveness updates
• Overall loading and speed improvements

Other

• Product team finalized roadmap for Q2
• Team meetups in Des Moines, Salt Lake City and Austin
• Roboflow hosted a recruiting event in Des Moines
• Visited a new customer's facility
• Updated core analytics tool to PostHog
• Added 2 new team members.
• Opened several positions (come help us democratize computer vision!)
• 26 bug fixes
• 36 discussions on the Roboflow Forum (opens in a new tab)
• 6 new Universe projects (opens in a new tab)
• 8 new blog posts
• 4 new YouTube videos (opens in a new tab)
• Featured in/by: TechCrunch (opens in a new tab), Ampera Racing (opens in a new tab), Nakashima (JP (opens in a new tab) ENG (opens in a new tab)), Queen's AutoDrive (opens in a new tab), AI for Wildlife (opens in a new tab).
• User projects: OAK + Roboflow Demo (opens in a new tab), SHEL5K: An Extended Dataset and Benchmarking for Safety Helmet Detection (opens in a new tab), Animal Behaviour and Speech Recognition (opens in a new tab), YOLOv5 Honey Bee Detection Model (opens in a new tab), Detection of Plant Diseases Using Leaf Images and Machine Learning (opens in a new tab), Identifying Fish Species (opens in a new tab), Fruit detection dataset (opens in a new tab), A Video Analytics System for Person Detection Combined with Edge Computing (opens in a new tab), Deep learning techniques to classify agricultural crops through UAV imagery: a review (opens in a new tab), IIOT Course (opens in a new tab), A visual approach towards forward collision warning for autonomous vehicles on Malaysian public roads (opens in a new tab), Deep learning techniques to classify agricultural crops through UAV imagery (opens in a new tab), Deep-Learning-Based Object Filtering According to Altitude for Improvement of Obstacle Recognition during Autonomous Flight (opens in a new tab), The practical guide for Object Detection with YOLOv5 algorithm (opens in a new tab), A Large-scale Dataset for Fine-grained and Domain-adaptive Recognition of Cereal Grains (opens in a new tab), Fast and Accurate Face Detector (opens in a new tab), Autonomous Mosquito Habitat Detection Using Satellite Imagery and Convolutional Neural Networks For Disease Risk Mapping (opens in a new tab), Development of an Aerial Fire Identification System Based on Visual Artificial Intelligence (opens in a new tab), Vehicle Detection for Vision-Based Intelligent Transportation Systems Using Convolutional Neural Network Algorithm (opens in a new tab), Trust-driven Reinforcement Selection Strategy for Federated Learning on IoT Devices (opens in a new tab)