Amortised Experimental Design and Parameter Estimation for User Models of Pointing

Abstract

Figure 1: An intelligent interactive system views a user through sensors that include a keyboard and mouse and sometimes, camera, microphone and/or eye-tracker. What is the best way to infer a model of the user from these data? In the figure gaze path is represented in red, mouse movements in black and text entered in blue. A camera captures the user's gaze direction.

Towards machines that understand people

“…Parameter estimation can be extremely expensive computationally, especially when full Bayesian posteriors are required. Here, an additional step in the workflow can involve training the model, in silico, on a distribution of possible user parameters (Moon et al 2022;Moon, Oulasvirta, and Lee 2023;Kwon et al 2020) and even training the cooperative AI to acquire user data that is maximally useful for parameter estimation (Ryan et al 2016;Foster et al 2019;Ivanova et al 2021;Keurulainen et al 2023).…”