IMAGE RECOGNITION AND SUGGESTION ON ELECTRONIC WHITEBOARDS
Publication Date: 2018-Apr-12
The IP.com Prior Art Database
A Machine Learning (ML) based image recognition system is provided that improves the readable output from whiteboard sessions. Thus, if a user wishes to draw an electronic circuit board on a whiteboard, after the "electronics" whiteboard template is selected, the system performs better at matching drawings to recognized electronic symbols. Moreover, a mechanism is also provided for image suggestion/autocorrect for whiteboard drawings and reference document suggestions during a discussion/meeting based on contextual data learned from the user, group, and company information. Image and document suggestions are derived from a ML driven learning model built based on the user's previous diagrams, projects, product documents, images, organizational wiki-portal, published product icons, etc. These ML driven, user context based image suggestions significantly improve collaboration user drawing experiences on electronic whiteboards. Similarly, the dynamic document references makes meeting more informative and productive for the users.