AI News, Walter S. Lasecki artificial intelligence

Real-time crowd labeling for deployable activity recognition

Systems that automatically recognize human activities offer the potential of timely, task-relevant information and support.

In this paper, we introduce Legion:AR, a system that provides robust, deployable activity recognition by supplementing existing recognition systems with on-demand, real-time activity identification using input from the crowd.

Legion:AR uses activity labels collected from crowd workers to train an automatic activity recognition system online to automatically recognize future occurrences.

CV

School in Logic &

ACM Conference on Fairness, Accountability, and Transparency (FAT*) 2020 Sociotechnical Considerations for Accessible Visualization Design Alan

Visualization &

Computer Graphics (InfoVis) 2020 Arboretum and Arbility: Improving Web Accessibility Through a Shared Browsing Architecture Steve

ACM Symposium on User Interface Software and Technology (UIST) 2018 Bolt: Instantaneous Crowdsourcing via Just-in-Time Training Alan

ACM Conference on Human Factors in Computing Systems (CHI) 2018 Graduate Research Fellowship 2019- National

Besmira Nushi

PUBLICATIONS

Metareasoning in Modular Software Systems: On-the-Fly Configuration using Reinforcement Learning with Rich Contextual Representations.

full technical report Crowd Access Path Optimization: Diversity Matters.

pdffull technical report When is A = B?

EATCS Bulletin 111 (2013) pdf Uncertain time-series similarity: Return to the basics.

11 (2012): 1662-1673.

pdf Similarity matching for uncertain time series: analytical and experimental comparison.