Application of deep learning (DL) methods on integrated photonics and/or quantum devices
Deep learning (DL), neural networks and artificial intelligence have already shown very promising results in our daily lives such as self-driving cars, image generation/recognition and text generation.
Not only that, but these techniques have also been applied to many scientific disciplines. To name a few, astrophysics , high energy physics , genetic engineering , new drug discovery  etc. Here, we are interested in applying DL algorithms for quantum physics related topics (please see project description).
For instance, when designing high fidelity quantum gate , deep learning techniques can be used to optimize the parameters of quantum gates, such as pulse shapes, gate durations, and amplitude profiles, to minimize errors and improve the overall fidelity of the gates. Deep learning algorithms can learn complex patterns and relationships from large datasets, which can aid in finding optimal gate parameters that may be challenging to determine analytically.
Deep learning can also be used to optimize the design parameters of photonic devices , such as geometries, material properties, and refractive indices, to achieve desired performance metrics. Deep learning algorithms can search through a large design space efficiently and provide insights into complex device physics, leading to novel device designs that may not be intuitive through traditional methods.
Common DL tools at your disposal would be feed forward network, recurrent neural network, variational auto encoder, reinforcement learning, generative adversarial network and many more.
Designing and implementing optimization and/or data analysis algorithms on integrated photonics and/or quantum devices based on diamond color centers.
In this project, you will explore one or more of the following topics (not exclusive) using DL methods:
- Optimizing the fidelity of quantum devices
- Optimizing the architecture/configuration of photonics components (waveguide/circuit/cavity)
- Reconstructing quantum states and/or quantum channels from measurement data
- Denoising quantum measurement data
Additionally, you may also conduct physical experiments based on the NV center platform (if applicable) using your algorithms as a proof-of-concept. Depending on your topic, it can be measuring the counts of quantum states for the reconstruction algorithms, or measuring the coupling efficiency of waveguide/circuit; quality factor for the cavity.
Students who are interested in this project are assumed to have basic understanding and/or experience of DL models.
 George, Daniel, and E. A. Huerta. "Deep neural networks to enable real-time multimessenger astrophysics." Physical Review D 97.4 (2018): 044039.
 Baldi, Pierre, Peter Sadowski, and Daniel Whiteson. "Searching for exotic particles in high-energy physics with deep learning." Nature communications 5.1 (2014): 4308.
 Wainberg, Michael, et al. "Deep learning in biomedicine." Nature biotechnology 36.9 (2018): 829- 838.
 Chen, Hongming, et al. "The rise of deep learning in drug discovery." Drug discovery today 23.6 (2018): 1241-1250.
 Zahedinejad, Ehsan, Joydip Ghosh, and Barry C. Sanders. "Designing high-fidelity single-shot threequbit gates: a machine-learning approach." Physical Review Applied 6.5 (2016): 054005.
 Jiang, Jiaqi, Mingkun Chen, and Jonathan A. Fan. "Deep neural networks for the evaluation and design of photonic devices." Nature Reviews Materials 6.8 (2021): 679-700.