PeopleofSuperintelligence

Researcher

Aliaksandra Shysheya

Also known as Sasha Shysheya

Aliaksandra Shysheya is a machine-learning researcher whose work spans neural avatars, few-shot learning, materials design and AI weather forecasting. Her early papers developed full-body avatars, personalized talking-head models and one-shot head synthesis, and she coauthored FiT for personalized and federated image classification. At Microsoft Research AI4Science, she contributed to MatterGen as an intern in 2023; she also coauthored a 2024 study of conditional diffusion for physical simulations and a 2025 Nature paper on MatterGen. As of October 2026, she is affiliated with Google DeepMind and coauthors WeatherNext 3, which uses satellite observations to produce hourly forecasts.

Works
15
Citations
1,481
h-index
6
Badges

Publication numbers from OpenAlex, as of 22 October 2026.

Timeline

  1. 2019

    Textured Neural Avatars

    In 2019, she coauthors Textured Neural Avatars, a full-body neural renderer that uses an explicit 2D texture map to create realistic views across body and camera poses.

  2. 2019

    Few-shot talking head models

    In 2019, she coauthors a talking-head system that meta-learns from videos and trains personalized models for previously unseen people from a few images.

  3. 2020

    One-shot head avatars

    In 2020, she coauthors a neural-rendering system that creates realistic head avatars from one photograph and reports faster inference at similar visual quality.

  4. 2022–2023

    FiT for few-shot learning

    She coauthors FiT in 2022, combining transfer learning and meta-learning for low-shot image classification, personalization and federated learning. The paper appears at ICLR in 2023 and achieves state-of-the-art accuracy on VTAB-1k with fewer than 1% of its updateable parameters.

  5. 2023

    Intern on MatterGen

    In December 2023, she contributes to the MatterGen collaboration as an intern at Microsoft Research AI4Science. The model generates stable, diverse inorganic materials and can be fine-tuned for target properties.

  6. 2024

    Diffusion for PDEs

    In 2024, she coauthors a comparative study of score-based diffusion models for forecasting and assimilating sparse observations in partial differential equations. The study introduces autoregressive sampling and a hybrid conditioning model.

  7. 2025

    MatterGen in Nature

    In 2025, she coauthors A generative model for inorganic materials design in Nature. The study describes MatterGen as a diffusion model for crystalline materials across the periodic table.

  8. As of October 2026

    Google DeepMind weather work

    As of October 2026, she is affiliated with Google DeepMind and coauthors WeatherNext 3. The model uses low-latency satellite data to generate new global probabilistic forecasts each hour.

Career

  1. NowGoogle DeepMind
  2. 2023MicrosoftIntern
  3. 2020–2024University of Cambridge

From the biography and OpenAlex affiliations (the years of papers that list the organization), so some years are approximate.

Most-cited works

  1. Few-Shot Adversarial Learning of Realistic Neural Talking Head Models2019 · IEEE/CVF International Conference on Computer Vision (ICCV) · 685 citationsWith Victor Lempitsky
  2. A generative model for inorganic materials design2025 · Nature · 509 citationsWith Andrew Fowler, Claudio Zeni, Daniel Zügner, Matthew Horton, Robert Pinsler and Ryota Tomioka
  3. Fast Bi-Layer Neural Synthesis of One-Shot Realistic Head Avatars2020 · Lecture notes in computer science · 143 citationsWith Victor Lempitsky
  4. Textured Neural Avatars2019 · IEEE Computer Society Conference on Computer Vision and Pattern Recognition · 111 citationsWith Victor Lempitsky
  5. Denoising Diffusion Probabilistic Models in Six Simple Steps2024 · 7 citationsWith Richard E. Turner

From OpenAlex, as of 21 October 2026.

  1. The median h-index is 48 at universities and 14 at frontier AI labsIncludes Aliaksandra Shysheya. The median h-index is 48 at universities and 14 at frontier AI labs. We compare publishing careers, recent citations, dual appointments and job moves across 2,420 university researchers and 308 frontier-lab researchers.
  2. 16% of AI career moves in the 2020s went from a university to a companyIncludes Aliaksandra Shysheya. We estimate that 16% of AI career moves in the 2020s went from a university to a company; 29% of moves starting at a university went to a company, and 76% ended the university role.
  3. Google had the most people shaping AI every year from 2014 to 2025Includes Google. Our estimates of which organizations employed the most people shaping AI each year from 1955 to 2025, tracing the shift from university labs to companies.
  4. Google has the highest combined h-index among AI organizationsIncludes Google. We estimate that Google holds the highest combined h-index at 8,016, while universities account for 73% of the deduplicated total across AI organizations.

Colleagues

People with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Aliaksandra Shysheya yet. Send us a free-licensed photo.

  1. Textured Neural Avatars openaccess.thecvf.com. Checked 20 October 2026.
  2. Few-Shot Adversarial Learning of Realistic Neural Talking Head Models arxiv.org. Checked 20 October 2026.
  3. Fast Bi-layer Neural Synthesis of One-Shot Realistic Head Avatars arxiv.org. Checked 20 October 2026.
  4. FiT: Parameter Efficient Few-shot Transfer Learning for Personalized and Federated Image Classification arxiv.org. Checked 20 October 2026.
  5. MatterGen: Property-guided materials design microsoft.com. Checked 20 October 2026.
  6. MatterGen: a generative model for inorganic materials design arxiv.org. Checked 20 October 2026.
  7. On conditional diffusion models for PDE simulations arxiv.org. Checked 20 October 2026.
  8. A generative model for inorganic materials design nature.com. Checked 20 October 2026.
  9. WeatherNext 3: Increasing resolution and performance of global weather models with raw observations ar5iv.labs.arxiv.org. Checked 20 October 2026.