PawelPierzchlewicz
A researcher and AI engineer working on probabilistic machine learning, motion modelling and inverse problems. CEO and co-founder of Animatica.ai, previously Head of Artificial Intelligence at TeaCode, where he led R&D teams on AI applications in animation and computer graphics. He began in science as a doctoral student with Professor Fabian Sinz at the University of Göttingen; his thesis, Deep Probabilistic Priors for Solving Inverse Problems, put deep probabilistic models such as normalising flows and diffusion models to work on inverse problems in neurobiology and human motion modelling. His research has been presented at NeurIPS and ICLR and published in journals including Cell Reports, and he is a co-author of Platypose, which gives calibrated multi-hypothesis 3D human pose estimates zero-shot, and of Energy Guided Diffusion, which uses diffusion models to generate neuron-stimulating images. Today he builds tools for generative character animation and works on getting AI into the pipelines film and games already run.





















