Dmitriy (Dima) Smirnov

Member of Technical Staff at OpenAI

San Francisco, CA

Dmitriy Smirnov

I am currently a research scientist at OpenAI, where I work on multimodal generation. Before joining OpenAI, I was a researcher at Reve and, previously, a research scientist at Netflix, working on machine learning and computer graphics for content creation. I earned my PhD in MIT CSAIL's Geometric Data Processing Group, advised by Prof. Justin Solomon. My research interests lie at the intersection of graphics, machine learning, computer vision, and geometry.

I have been fortunate to intern at Pixar Research, Adobe Research, Harvey Mudd College, the Lawrence Berkeley National Laboratory, and Google Maps. Before MIT, I earned my B.A. in Math and Computer Science from Pomona College, where I was advised by Prof. Vin de Silva.

In my spare time, I enjoy photography, cooking, and mixology.

Product

Reve regional color control with the color picker open

Reve

Controllable image generation.

Two participants in an outdoor group embrace in Trust Me: The False Prophet

Netflix

AI face anonymization for Trust Me: The False Prophet.

Research

2026

DiffHDR: Re-exposing LDR Videos with Video Diffusion Models
@inproceedings{yu2026diffhdr,
  title={DiffHDR: Re-exposing LDR Videos with Video Diffusion Models},
  author={Yu, Zhengming and Ma, Li and He, Mingming and Isikdogan, Leo and Xu, Yuancheng and Smirnov, Dmitriy and Salamanca, Pablo and Mi, Dao and Delgado, Pablo and Yu, Ning and Philip, Julien and Li, Xin and Wang, Wenping and Debevec, Paul},
  year={2026},
  booktitle={Proceedings of the European Conference on Computer Vision (ECCV)}
}
Less is More: Data-Efficient Adaptation for Controllable Text-to-Video Generation
@inproceedings{cheng2026lessismore,
  title={Less is More: Data-Efficient Adaptation for Controllable Text-to-Video Generation},
  author={Cheng, Shihan and Kulkarni, Nilesh and Hyde, David and Smirnov, Dmitriy},
  year={2026},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}
}

2025

Infinite-Resolution Integral Noise Warping for Diffusion Models
@inproceedings{deng2024infiniteresolution,
  title={Infinite-Resolution Integral Noise Warping for Diffusion Models},
  author={Deng, Yitong and Lin, Winnie and Li, Lingxiao and Smirnov, Dmitriy and Burgert, Ryan and Yu, Ning and Dedun, Vincent and Taghavi, Mohammed H.},
  year={2025},
  booktitle={Proceedings of the International Conference on Learning Representations (ICLR)}
}

2024

2023

2022

Deep Learning on Geometry Representations

Deep Learning on Geometry Representations

Dmitriy Smirnov

PhD Dissertation, MIT Department of Electrical Engineering and Computer Science, 2022

@phdthesis{smirnov2022deep,
  title={Deep Learning on Geometry Representations},
  school={Massachusetts Institute of Technology},
  author={Smirnov, Dmitriy},
  year={2022},
}
DeepCurrents: Learning Implicit Representations of Shapes with Boundaries
@inproceedings{palmer2022deepcurrents,
  title={{DeepCurrents}: Learning Implicit Representations of Shapes with Boundaries},
  author={Palmer, David and Smirnov, Dmitriy and Wang, Stephanie and Chern, Albert and Solomon, Justin},
  year={2022},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}
}

2021

MarioNette: Self-Supervised Sprite Learning
@inproceedings{smirnov2021marionette,
  title={{MarioNette}: Self-Supervised Sprite Learning},
  author={Smirnov, Dmitriy and Gharbi, Micha\"el and Fisher, Matthew and Guizilini, Vitor and Efros, Alexei A. and Solomon, Justin},
  year={2021},
  booktitle={Advances in Neural Information Processing Systems (NeurIPS)}
}
Interactive All-Hex Meshing via Cuboid Decomposition
@article{li2021hex,
  title={Interactive All-Hex Meshing via Cuboid Decomposition},
  author={Li, Lingxiao and Zhang, Paul and Smirnov, Dmitriy and Abulnaga, Mazdak and Solomon, Justin},
  month={December},
  year={2021},
  journal={ACM Transactions on Graphics (TOG)},
  publisher={ACM},
  volume={40},
  number={6},
  pages={256:1--256:17}
}
Polygonal Building Segmentation by Frame Field Learning

Polygonal Building Segmentation by Frame Field Learning

Nicolas Girard, Dmitriy Smirnov, Justin Solomon, Yuliya Tarabalka

CVPR 2021, online (Oral Presentation, Best Paper Finalist)

IGARSS 2020, online (Oral Presentation)

@inproceedings{girard2021pbs,
  title={Polygonal Building Segmentation by Frame Field Learning},
  author={Girard, Nicolas and Smirnov, Dmitriy and Solomon, Justin and Tarabalka, Yuliya},
  year={2021},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}
}

2020

Deep Parametric Shape Predictions using Distance Fields
@inproceedings{smirnov2020dps,
  title={Deep Parametric Shape Predictions using Distance Fields},
  author={Smirnov, Dmitriy and Fisher, Matthew and Kim, Vladimir G. and Zhang, Richard and Solomon, Justin},
  year={2020},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}
}

Earlier work

Visualizing Scissors Congruence
@inproceedings{devadoss2016scissors,
  title={Visualizing Scissors Congruence},
  author={Devadoss, Satyan L. and Epstein, Ziv and Smirnov, Dmitriy},
  year={2016},
  booktitle={32nd International Symposium on Computational Geometry (SoCG 2016)},
  series={Leibniz International Proceedings in Informatics (LIPIcs)},
  volume={51},
  pages={66:1--66:3},
  doi={10.4230/LIPIcs.SoCG.2016.66}
}
DTL-RnB: Algorithms and Tools for Summarizing the Space of DTL Reconciliations
@article{ma2018dtlrnb,
  title={{DTL-RnB}: Algorithms and Tools for Summarizing the Space of {DTL} Reconciliations},
  author={Ma, Weiyun and Smirnov, Dmitriy and Forman, Juliet and Schweickart, Annalise and Slocum, Carter and Srinivasan, Srinidhi and Libeskind-Hadas, Ran},
  year={2018},
  journal={IEEE/ACM Transactions on Computational Biology and Bioinformatics},
  volume={15},
  number={2},
  pages={411--421},
  doi={10.1109/TCBB.2016.2537319}
}