Jameson TooleinHeartbeatOptimizing a HuggingFace Transformer Model for Toxic Speech DetectionAs someone who has designed neural networks destined for real-time use on mobile devices, state-of-the-art NLP models like transformers…May 12, 2021May 12, 2021
Jameson TooleSnapchat Lens Recipe: “Clone”A quick look at how we built our SnapML-powered Lens that lets you instantly crop, clone, and manipulate objects in both 2D and 3DMar 4, 2021Mar 4, 2021
Jameson TooleinHeartbeatImplementing Conway’s Game of Life in Lens StudioHow to use Game of Life theory to create a snapchat lensFeb 1, 2021Feb 1, 2021
Jameson TooleBuilding an on-device face mask detector with Fritz AI StudioRecognizing whether or not a user is wearing a face mask in real-time with mobile machine learningMay 29, 2020May 29, 2020
Jameson TooleAdding custom object detection to an iOS app with Turi Create and Fritz AILocate and identify objects on-deviceNov 13, 2019Nov 13, 2019
Jameson TooleinHeartbeatDeep learning has a size problemShifting from state-of-the-art accuracy to state-of-the-art efficiencyNov 5, 20191Nov 5, 20191
Jameson TooleinAndroidPubCreating a Pet Sticker App on Android with Fritz Pet SegmentationTurn dogs and cats into stickers to use in your favorite messaging apps.Jul 15, 2019Jul 15, 2019
Jameson TooleOrganizing mobile machine learning projects with the Fritz CLIA suite of tools designed to streamline mobile ML workflowsJul 10, 2019Jul 10, 2019
Jameson TooleinBetter ProgrammingCreating a Pet Sticker App on iOS with Fritz Pet SegmentationTurn dogs and cats into stickers to use in your favorite messaging apps.Jun 20, 2019Jun 20, 2019
Jameson TooleinHeartbeatWhat’s new in Core ML 3On-device training, 100+ model layer types, smoother conversion, and more.Jun 11, 20192Jun 11, 20192