Neural networks trained on datasets such as ImageNet have led to major advances in visual object classification. One obstacle that prevents networks from reasoning more deeply about complex scenes and situations, and from integrating visual knowledge with natural language, like humans do, is their lack of common sense knowledge about the physical world. Videos, unlike still images, contain a wealth of detailed information about the physical world. However, most labelled video datasets represent high-level concepts rather than detailed physical aspects about actions and scenes. In this work, we describe our ongoing collection of the "something-something" database of video prediction tasks whose solutions require a common sense understanding of the depicted situation. The database currently contains more than 100,000 videos across 174 classes, which are defined as caption-templates. We also describe the challenges in crowd-sourcing this data at scale.
Abstract. We present a software tool platform which facilitates security and performance analysis of systems which starts and ends with UML model descriptions. A UML project is presented to the platform for analysis, formal content is extracted in the form of process calculi descriptions, analysed with the analysers of the calculi, and the results of the analysis are reflected back into a modified version of the input UML model. The design platform supporting the methodology, Choreographer, interoperates with state-of-the-art UML modelling tools. We illustrate the approach with a well known protocol and report on the experience of industrial users who have applied Choreographer in their development work.
We present a novel method of assuring security and performance demands on systems based on automated analysis of UML model descriptions. Analysable content is extracted from the UML models in the form of process calculus descriptions. These are analysed to provide strong guarantees of satisfactory security and performance. The results are reflected back in the form of a modified version of the UML model which highlights points of the design which can give rise to operational difficulties. A design platform supporting the methodology, Choreographer, interoperates with state-of-the-art UML modelling tools such as Poseidon. We illustrate the approach on an example provided by our industrial partner.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.