Avatars Grow Legs: Generating Smooth Human Motion from Sparse Tracking Inputs with Diffusion Model
Yuming Du, Robin Kips, Albert Pumarola, Sebastian Starke, Ali Thabet, Artsiom Sanakoyeu
International Symposium on Mixed and Augmented Reality (ISMAR)
Supporting refractive correction in head-mounted AR systems is central for providing accessibility across a diverse population. Importantly, matching the visual experience between a user's traditional ophthalmic lenses and their AR devices is critical for a seamless and comfortable user experience. In terms of geometric distortion, this can be challenging with the addition of optical and non-optical elements that accompany the AR experience. We developed an analytical model that quantifies the binocular aspects of optically induced geometric distortions with an aim to develop perceptually based metrics that can predict the user's experience with different spectacle lens designs. We anchored this model against empirical data collected from a small study where we systematically varied the front surface curvature of a user's habitual refractive correction. Importantly, the models we derived explained a significant amount of variance (r2 = 0.46 to 0.92) in users’ reported visual comfort. These results support the value of quantifying the binocular aspects of the visual experience in see-through optical systems and lay a foundation for a user-centric, quantitative system that can be used to evaluate optical lens designs in both the ophthalmic and near-eye display industries.
Yuming Du, Robin Kips, Albert Pumarola, Sebastian Starke, Ali Thabet, Artsiom Sanakoyeu
Lisa Rivalin, Andrew Grier, Tobias Tiecke, Chi Zhou, Doris Gao, Prakriti Choudhury, John Fabian
Nadia Alshahwan, Mark Harman, Alexandru Marginean