No abstract

Identifying what people do in the home can both inform ubiquitous computing application design decisions and provide training data to the machine learning algorithms used in their implementation. This paper describes an unsupervised technique in which contextual information gathered by ubiquitous sensors is used to help users label a multitude of anonymous activity episodes. This context-aware recognition survey is a game-like computer program in which users attempt to correctly guess which activity is happening after seeing a series of symbolic images that represent sensor values generated during the activity. We report a user study of the system, focusing on how well subjects were able to recognize their own activities, the activities of others, and counterfeits that did not correspond to any activity.

Object geo-localization from images is crucial to many applications such as land surveying, self-driving, and asset management. Current visual object geo-localization algorithms suffer from hardware limitations and impractical assumptions limiting their usability in real-world applications. Most of the current methods assume object sparsity, the presence of objects in at least two frames, and most importantly they only support a single class of objects. In this paper, we present a novel two-stage technique that detects and geo-localizes dense, multi-class objects such as traffic signs from street videos. Our algorithm is able to handle low frame rate inputs in which objects might be missing in one or more frames. We propose a detector that is not only able to detect objects in images, but also predicts a positional offset for each object relative to the camera GPS location. We also propose a novel tracker algorithm that is able to track a large number of multi-class objects. Many current geo-localization datasets require specialized hardware, suffer from idealized assumptions not representative of reality, and are often not publicly available. In this paper, we propose a public dataset called ARTSv2, which is an extension of ARTS dataset that covers a diverse set of roads in widely varying environments to ensure it is representative of real-world scenarios. Our dataset will both support future research and provide a crucial benchmark for the field.

From the intersection of computational science and technological speculation, with boundaries limited only by our ability to imagine what could be. When machines are in the natural world, what in the world is still unnatural?

Engineers and researchers, particularly in the field of robotics and human-computer interaction, are often inspired by science fiction futures depicted in novels, on television, and in the movies. For example, Honda's Asimo humanoid robot is said to have been directly inspired by the Astroboy manga series.In turn, public perception of science is also shaped by science fiction. For better or worse, broad technological expectations of the future (aesthetic and otherwise) are largely set by exposure to science fiction in popular culture. These depictions have a direct impact on attitudes toward new technology.We review some common tropes of science fiction (including the idea of the "singularity" and killer robots) and examine why certain archetypes might persist while others fall by the wayside. From the perspective of a scientistturned-sci-fi-author, we discuss factors that go into the creation of science fiction and how these factors may or may not correspond to the needs and wants of the actual science community.Exposure to science fiction influences scientists and the general public, both to build and adopt new technologies. The inextricable link between science and science fiction helps to determine how and when those futures arrive. ACM ClassificationA.1 General Literature, INTRODUCTORY AND SURVEY