Scratchpad

Digital look-alikes, lighting capture, simulation and relighting

• 'High-fidelity facial reflectance and geometry inference from an unconstrained image' at dl.acm.org, a w:deep learning-based technique described in a paper presented at the w:ACM Transactions on Graphics in July 2018
• 'Deep reflectance fields: high-quality facial reflectance field inference from color gradient illumination' at dl.acm.org View video of the presentation of the work presented at the w:ACM Transactions on Graphics in July 2019
• 'Deep view synthesis from sparse photometric images' at dl.acm.org View video of the presentation of the work presented at the w:ACM Transactions on Graphics in July 2019
• 'Depth estimation meets inverse rendering for single image novel view synthesis' at dl.acm.org published in CVMP '19: European Conference on Visual Media Production in December 2019
• 'Computational Methods for Capture and Reproduction of Photorealistic Surface Appearance' at aaltodoc.aalto.fi, a 2016 doctoral dissertation by Miika Aittala at the w:Aalto University
• 'Local light field fusion: practical view synthesis with prescriptive sampling guidelines' at dl.acm.org presented at the w:ACM Transactions on Graphics in July 2019
• 'Deep view synthesis from sparse photometric images' at dl.acm.org presented at the w:ACM Transactions on Graphics in July 2019

w:Maeil Broadcasting Network in South Korea published an AI news anchor in November 2020 made in conjunction with MoneyBrain. You can enable YouTube auto-generated English subtitles by enabling closed captioning (cc) and then clicking on the setting icon on the video control bar.

Digital sound-alikes

• 'Generative Adversarial Networks for Speech Synthesis' at aaltodoc.aalto.fi, a master's thesis by Kasperi Palkama from June 2020 in w:Aalto University