Mobile Context-Aware Inference Toolbox and Service [Report]

NESL Technical Report #: 2013-3-3


Abstract: Emerging continuous sensing applications pose new challenges on smartphone’s software design. In this paper, we identified 6 classifiers and 9 sensors that can cover 70% of recent inference models for continuous sensing applications. we will describe the design of Mobile Context-Aware Inference Toolbox and Service. The design has two components: a toolbox that provide programmers with commonly used building blocks so they can develop inference models easily; a centralized service that can achieve CPU time and power saving internally by sharing feature extraction result among classifiers as well as sharing classifier inference result among applications.

Publication Forum: MS Project Report, UCLA Electrical Engineering Department

Page (Start): 1

Page (End): 10

Page (Count): 10

Date: 2013-03-01

Place: Los Angeles, CA

Public Document?: Yes

NESL Document?: Yes

Document category: Report