Cecilia Curreli

Profile picture
I am pursuing my PhD at the Technical University of Munich with Prof. Cremers - Check out the awesome work of the Computer Vision Group! My research focuses on generative models for 4D computer vision (diffusion) and 3D human motion prediction. What intrigues me is the probabilistic nature of reality and what we still can learn from the temporal dimension.

I am currently in a Research Internship at Google working with Osman Ulusoy.Last spring I joined the Scene Representation group of Prof. Vincent Sitzmann at CSAIL, MIT, as a visiting researcher. I am a junior member of the Munich Center of Machine Learning, one of six national AI Competence Centers in Germany.

Before being part of CVG, I joined the Lab of Prof. Shin'ichi Satoh at the University of Tokyo and National Institute of Informatics. I completed my Master's thesis with Prof. Matthias Nießner and my Bachelor's degree with Prof. Leal Taixè. With Prof. Steffen Glaser (Munich Quantum Valley initiative and Bavarian NMR Center), I pursued research related to sequences of quantum particles. In 2018, I was part of the Chinese Academy of Sciences in Suzhou as an intern with Prof. Xiaodong.

Updates

  • July 2026 Our paper EquiFusion: Kinematics-Agnostic Human Motion Prediction via Equivariant Latent Diffusion was accepted to ECCV 2026
  • July 2026 Our paper Evaluating Super-Resolution Models for Real-World Sentinel-2 Applications: A Case Study was accepted and accepted as an oral presentation to ISPRS Annals 2026 in Toronto
  • May 2026 I joined Google for a research internship in Seattle, with two great hosts Osman Ulusoy and Connor Lee at Google Beam.
  • May 2026 Our paper Benchmarking Single-Step Inpainting Methods for Multi-Object 3D Gaussian Splatting Scenes was accepted as an extended abstract at CVEU Workshop at CVPR 2026
  • May 2025 I joined the MCML delegation to present our research at NYU, Cornell, Cornell Tech, Harvard, and MIT.
  • Apr 2025 Our paper Prototype Augmented Hypernetworks was accepted to CVPR 2025 Workshop
  • Feb 2025 Nonisotropic Gaussian Diffusion was accepted to CVPR 2025
  • Jan 2025 Pre-print of Nonisotropic Gaussian Diffusion for Realistic 3D Human Motion Prediction is online
  • Dec 2024 Our paper ZDySS - Zero-Shot Dynamic Scene Stylization using Gaussian Splatting is online
  • Aug 2024 Our paper Gaussian Splatting in Style was accepted to GCPR 2024
  • Sep 2023 I joined the Computer Vision Group at TUM as a PhD candidate with Prof. Daniel Cremers

Research

See my full list of publications.