Citation and History
If you use DeepLens in your research, please cite the relevant papers.
Primary Citation
Curriculum Learning for Ab Initio Deep Learned Refractive Optics
@article{yang2024curriculum,
title={Curriculum learning for ab initio deep learned refractive optics},
author={Yang, Xinge and Fu, Qiang and Heidrich, Wolfgang},
journal={Nature Communications},
volume={15},
number={1},
pages={6572},
year={2024},
publisher={Nature Publishing Group UK London}
}
Paper link: https://www.nature.com/articles/s41467-024-50835-7
History of DeepLens
Evolution Timeline
2021: Early Development
Initial concept and prototyping
Basic ray tracing framework
Differentiable optimization proof-of-concept
2022: Core Features
Comprehensive optical surface library
Wave optics integration
Sensor and ISP simulation
First automated lens designs
2023: Advanced Capabilities
Aberration-aware imaging (TPAMI 2023)
Neural surrogate models (PSFNet)
End-to-end optimization framework
Multi-wavelength support
2024: Public Release
2025: Ongoing Development
Key Milestones
Date |
Milestone |
2021 |
Project initiated at KAUST |
2022 |
First fully automated lens design from scratch |
2023 Jun |
TPAMI paper on aberration-aware depth estimation |
2024 Jul |
Nature Communications paper published |
2024 Aug |
Open-source release on GitHub |
2024 Dec |
SIGGRAPH Asia paper on hybrid lenses |
2025 |
Community edition with extended features |
Contributors
Core Team
Xinge Yang - Project Lead, Main Developer
Qiang Fu - Optical Design Consultant
Wolfgang Heidrich - Principal Investigator
Acknowledgments
Funding
This project was supported by:
King Abdullah University of Science and Technology (KAUST)
Visual Computing Center (VCC)
Computational Imaging Lab
Collaborations
We thank our collaborators:
Awards and Recognition
Nature Communications - Featured article (2024)
SIGGRAPH Asia - Technical paper (2024)
TPAMI - Journal publication (2023)
Research Using DeepLens
If you have published research using DeepLens, let us know! We’d love to feature it here.
Example applications:
Automated lens design for imaging systems
Computational photography research
Virtual reality optics
Smartphone camera optimization
Microscopy and scientific imaging
Automotive camera systems
License
DeepLens is released under the Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0).
What this means:
✓ Use for academic and research purposes
✓ Modify and build upon the work
✓ Share with attribution
✗ Commercial use without permission
For commercial licensing, please contact the authors.
Future Directions
Research Directions
Quantum optical systems
Freeform optics
Holographic displays
Neuromorphic imaging
Computational microscopy