Zensors: Adaptive, Rapidly Deployable, Human-Intelligent Sensor Feeds

摘要: The promise of "smart" homes, workplaces, schools, and other environments has long been championed. Unattractive, however, the cost to run wires install sensors. More critically, raw sensor data tends not align with types questions humans wish ask, e.g., do I need restock my pantry? Although techniques like computer vision can answer some these questions, it requires significant effort build train appropriate classifiers. Even then, systems are often brittle, limited ability handle new or unexpected situations, including being repositioned environmental changes (e.g., lighting, furniture, seasons). We propose Zensors, a sensing approach that fuses real-time human intelligence from online crowd workers automatic approaches provide robust, adaptive, readily deployable intelligent With users go question live feed in less than 60 seconds. Through our API, Zensors enable variety rich end-user applications moves us closer responsive, environments.