nerd thread
I gave a @ARealityEvent talk on why human expects dynamic output when we give immersive computers our context. Proceduralism + ML are needed to meet those expectations.

Some of you asked for the resources and research I mentioned, so here we go... #AWE2020

Generating layouts with the consideration of architectural "style"

Stanislas Chaillou. 2019. “AI + Architecture | Towards a New Approach” Harvard Graduate School of Design. https://towardsdatascience.com/ai-architecture-f9d78c6958e0

Estimating pose from videos created with widely accessible devices (your phone)

Xue Bin Peng and Angjoo Kanazawa. 2018. “Learning Acrobatics by Watching YouTube” UC Berkeley https://bair.berkeley.edu/blog/2018/10/09/sfv/

working together with Jun #adoberesearch on how to incorporate spatial semantics in the scene to emulate realistic and probable animations.

Sebastian Starke, He Zhang, Taku Komura, Jun Saito. 2019 “Neural state machine for character-scene interactions”

https://dl.acm.org/doi/10.1145/3355089.3356505

Together with J from #adoberesearch, we look to leverage genetic algorithm based on different input fidelity.

J Eisenmann, M Lewis, R Parent. 2013. "Inverse Mapping with Sensitivity Analysis for Partial Selection in Interactive Evolution" https://www.researchgate.net/publication/262399975_Inverse_Mapping_with_Sensitivity_Analysis_for_Partial_Selection_in_Interactive_Evolution