Manually labeling video datasets for segmentation tasks is extremely time consuming. In this paper, we introduce ScribbleBox, a novel interactive framework for annotating object instances with masks in videos. In particular, we split annotation into two steps: annotating objects with tracked boxes, and labeling masks inside these tracks. We introduce automation and interaction in both steps. Box tracks are annotated efficiently by approximating the trajectory using a parametric curve with a small number of control points which the annotator can interactively correct. Our approach tolerates a modest amount of noise in the box placements, thus typically only a few clicks are needed to annotate tracked boxes to a sufficient accuracy. Segmentation masks are corrected via scribbles which are efficiently propagated through time. We show significant performance gains in annotation efficiency over past work. We show that our ScribbleBox approach reaches 88.92% J&F on DAVIS2017 with 9.14 clicks per box track, and 4 frames of scribble annotation.
Curious how deep learning solutions are affecting your industry? On Wednesday, March 18th, join Upward, Connecticut Center for Advanced Technology (CCAT), and Connecticut Technology Council (CTC) for Hartford's first "teCTalk" focused on artificial intelligence/machine learning. Learn how machines are being taught to interact with the organic world around them and how this smart technology is working to elevate modern business. An esteemed panel of AI/ML experts across industries, including Upward Citizens GalaxE Solutions, VAANGO, and Saya Life, will navigate participants through the complex topic of deep learning. This event is designed to be interactive: pose your questions to the experts and engage with others in the crowd!
The Microsoft partner and customer network brings in more than $100 billion in revenue each year, most of the company's earnings. Keeping tabs on the millions of annual transactions is no small task--just ask Shashi Lanka Venkata and Mark Anderson, two company employees who are leading a bid to automate what historically has been a painstakingly manual revenue transaction process. "We support close to 50 million platform actions per day," says Venkata, a principal group engineering manager in Microsoft Core Services Engineering and Operations (CSEO). "For a quarter-end or a month-end, it can double. That's a lot, especially when there cannot be any mistakes and every transaction must be processed in 24 hours. To wrangle that high-stakes volume, Venkata and Anderson, a director on Microsoft's Business Operations team, teamed up to expand the capabilities of Customer Obsessed Solution Management and Incident Care (COSMIC), a Dynamics 365 application built to help automate Microsoft's ...
A Toronto-based content marketing and live publishing SaaS platform is boosting its big data capabilities with the acquisition of a New York City-based influence analytics marketing platform. ScribbleLive announced its acquisition on Wednesday of Appinions Inc., developers of an influence analytics marketing platform that is built on natural language processing and social network analysis technology. The vendor says combining ScribbleLive's data centric content workflow and publishing platform with Appinions' deep data technology will create an even more powerful content marketing platform for data-driven marketers. Toronto-based web platform creators Scribble Technologies Inc. has acquired another well-known live publishing platform in CoveritLive from California-based Demand Media. "By enabling over 500 brands such as Red Bull, PwC and Oracle to serve up the right content at the right time in the right place online, ScribbleLive has been validated as the platform for planning, creating and distributing content," said Vince Mifsud, CEO of ScribbleLive, in a statement.
Apple unveiled a range of new features for iPadOS 14 at its keynote presentation at the Worldwide Developers Conference. The biggest change in the new tablet operating system focus on Apple Pencil Support, part of the company's continuing effort to make handwritten inputs as effective as a keyboard. To that end, Apple has developed'Scribble,' a new iPadOS feature that will allow users to handwrite into every text field offered by any app on the system. Apple's latest update to iPadOS 14 will make handwriting through the Apple Pencil a major focus, part of the company's efforts to make handwritten inputs as effective as a keyboard The feature will automatically convert handwriting to typed lettering, and will also recognize Chinese characters in addition to the standard English alphabet. Scribble will let users switch back and forth between handwritten English and Chinese within the same document and even in the same line, and will automatically translate both English lettering and Chinese characters into clean type.