It's been a whirlwind of a month at Roboflow with lots of changes, updates, and expansion going on behind the scenes. 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.
Roboflow Organize
- Released "Isolate Objects" preprocessing step to extract bounding boxes from object detection datasets for use in single-class classification
- Added ability to add image annotations via REST API
- Improved Scale.ai annotation support
- Added a global in-app announcement ability
Roboflow Train and Deploy
- Launched Enterprise GPU accelerated inference container
- Added image resize to sample web app code
- Improved inference usage logs and reporting
- Created offline mode example repo
- Introduced inference server benchmark repo
- Version controlled docs
Roboflow Universe
- Improved tooling for finding and featuring projects
- Speed improvements across the backend infrastructure
- Highlighted 16 new projects
Other
- Doubled the size of the team (come help us democratize computer vision!)
- Had full-team on-site in Phoenix, AZ
- Updated all training notebooks to use our PIP package & API
- Refactored blog theme
- 20 bug fixes
- 9 blog posts
- 6 YouTube videos
- Presentations: Kimo CV
- Featured in/at: MarketWatch, Yahoo News, The Tech Tribune
- User projects: coral reef restoration, football analysis, mold detection, environment exploration, plastic waste sorting, object tracking, hydroponic farming, face detection, drone robotics challenge, plant breeding, self driving car