Yuheng Chen's Portfolio

Machine learning for device optimization

(collaborated with Microsoft, QuEra, and Oak Ridge National Lab)

Skills:

Python

PyTorch

Photonics Simulation (Comsol & Lumerical)

Deep Learning

Quantum Computation

Large Language Model (LLM)

We present using machine learning/quantum computing forcharacterization, fabrication, and inverse design of device applications, such as adjoint simulation, generative model-assisted design, hybrid quantum-classical optimization, LLM model-embedded multimodal training.

Project 1:Advancing photonic design with topological latent diffusion generative model

• Developed topology optimization (TO) based deep generative model: Topological Latent Diffusion Model (TLDM), to realize high-quality inverse design.
• Applied efficiency prediction model-embedded conditional U-net and demonstrated substantial efficiency improvement compared with state-of-the-art generative model benchmarks.

Project 2:Variational neural annealing for latent polynomial unconstrained binary optimization (PUBO) in device design

• Mapped device optimization problem into latent PUBO energy model to enforce the combinatorial optimization problem to the data optimization problem.
• Introduced variational neural annealing implemented through recurrent neural networks (RNNs) to solve PUBO, significantly outperformed simulated annealing and quantum annealing on sampling time and device efficiency.

Project 3:Multimodal model for prompt-guided integrated photonics design

• Combined with ChatGPT API with the packaged trained model to realize an interactive LLM-empowered prompt-guided photonics design interface.

• Utilized stable diffusion model for a device feature description text-device topology image multimodal dual-training.

Publications:

1. M. Bezick†, Y. Chen†, B. Wilson, A. V. Kildishev, V. M. Shalaev, and A. Boltasseva, 'Latent diffusion models for global optimization in inverse design', arxiv, Nature Communications,in review.
2. M. Bezick, B. Wilson, V. Iyer, Y. Chen, V. M. Shalaev, S. Kais, A. Boltasseva, and B. Lackey, 'Pearson-correlated variational neural annealing for latent PUBO optimization', arxiv, Advanced Optical Materials,in review.
3. Y. Chen, M. Bezick, B. Wilson, O. Yesilyurt, A. V. Kildishev, A. Boltasseva, and V. M. Shalaev, 'Advancing photonic design with topological latent diffusion generative model', Optica Frontiers in Optics + Laser Science Conference (2024).
4. B. Wilson, Y. Chen, V. Shalaev, A. Kildishev, and A. Boltasseva, 'Advancing nano- and quantum photonics with machine learning', International Conference on Metamaterials, Photonic Crystals and Plasmonics (2024).
5. B. Wilson, Y. Chen, S. Kais, A. Kildishev, V. Shalaev, and A. Boltasseva, 'Empowering quantum 2.0 devices and approaches with machine learning', Optica Quantum 2.0 Conference and Exhibition, QTu2A.13 (2022).[Paper]
6. Y. Chen, B. Wilson, A. Kildishev, V. Shalaev, and A. Boltasseva, 'Generative models for photonics device design and optimization',in preparation.

Machine learning for semiconductor

Skills:

Python

PyTorch

Computer Vision

Deep Learning

Image Segmentation

Data Augmentation

The global chip industry is grappling with dual challenges: a profound shortage of new chips and a surge of counterfeit chips valued at $75 billion, introducing substantial risks of malfunction and unwanted surveillance. To counteract this, we propose optical anti-counterfeiting methods for semiconductor devices that is robust under adversarial tampering features.

Project 1:Authentication in chip encryption through deep engine-based processing of tampered optical responses

• Designed a RAPTOR (Residual, Attention-based Processing of Tampered Optical Response) discriminator for identifying adversarial tampering of an optical, physical unclonable function based on a random array of gold nanoProjecticles embedded in semiconductor packaging.

• Extracted features of gold nanoProjecticles from 1000 dark-field images in just 27 ms and verified their authenticity using RAPTOR in 80 ms with 97.6% accuracy under difficult adversarial tampering conditions.

Publications:

1. B. Wilson†, Y. Chen†, D. K. Singh, R. Ojha, J. Pottle, M. Bezick, A. Boltasseva, V. M. Shalaev, and A. V. Kildishev, 'Authentication through residual attention-based processing of tampered optical responses', Advanced Photonics, 6(5), 056002-056002 (2024).[Paper]
2. B. Wilson†, Y. Chen†, D. K. Singh, R. Ojha, J. Pottle, M. Bezick, V. M. Shalaev, A. Boltasseva, and A. V. Kildishev, 'Machine learning assisted optical authentication of chip tampering', SPIE Optics + Photonics, Metamaterials, Metadevices, and Metasystems Conference, 13113-16 (2024).
3. SPIE featured news on our work: 'AI-powered optical detection to thwart counterfeit chips, researchers developed a robust optical anticounterfeit method for semiconductor devices'.[News]

Nanophotonics and Nanofabrication (With USTC and ASU)

Skills:

Photonics Simulation (Comsol & Lumerical)

Nanofabrication (cleanroom experience in process design & lithographing & etching & characterization)

Python

Matlab

We demonstrate novel nanofabrication method for light-matter plasmonic enhancement based on 2D materials. Also explore metasurface designs for efficient solar energy conversion.

Project 1:Integrated 2D semiconductor light-emitting devices with plasmonic nanostructures

• Realized first experimental transfer and emission characterization of 2D TMDs (Transition-metal dichalcogenides) on plasmonic nano-terrace morphology.
• Generated 12-fold light emission enhancement with flexible manipulation feature, LSPR (Localized surface plasmon resonance) enhancement mechanism verified through FDTD simulation.

Project 2:Self-organized lithography-free nanodevice fabrication with tunable optical anisotropy

• Implemented lithography-free nanofabrication method as team leader, realizing 3-fold aspect ratio promotion in self-organized metal co-deposition ion etching.
• Demonstrated outstanding tunable optical anisotropy feature in polarization, fitting well with FDTD/RCWA simulation.

Publications:

1. L. Mascaretti, Y. Chen, O. Henrotte, O. Yesilyurt, V. M. Shalaev, A. Naldoni, and A. Boltasseva, 'Designing metasurface for efficient solar energy conversion', ACS Photonics, 10(12), 4079-4103 (2023).[Paper]
2. Y. Chen, H. Li, M. Blei, M. Cai, H. Zang, Y. Lu, S. Tongay, and Y. Liu, 'Monolayer excitonic semiconductors integrated with Au quasi-periodic nanoterrace morphology on fused silica substrates for light-emitting devices', ACS Applied Nano Materials, 4, 84-93 (2021).[Paper]
3. Y. Chen, M. Cai, H. Zang, H. Chen, S. Kroker, Y. Lu, Y. Liu, F. Frost, and Y. Hong, 'Optical anisotropy of self-organized gold quasi-blazed nanostructures based on a broad ion beam', Applied Optics, 60, 505-512 (2021).[Paper]
4. Y. Chen, M. Cai, K. Qiu, and Y. Hong, 'Optical anisotropy of metal nanowire arrays on fused silica surface', Proceedings of SPIE, 114271N (2020).[Paper]
5. M. Cai, Z. Chen, Y. Chen, K. Qiu, X. Xu, and Y. Hong, 'Design of near-infrared resonance antenna array filters in termophotovoltaic application', Proceedings of SPIE, 114274E (2020).[Paper]

Talks and Presentations

1. 'Advancing photonic design with topological latent diffusion generative model', Optica Frontiers in Optics + Laser Science Conference (2024).
2. 'Machine learning assisted optical authentication of chip tampering', SPIE Optics + Photonics, Metamaterials, Metadevices, and Metasystems Conference (2024).
3. 'Topological latent diffusion model-assisted meta-atom design', Gordon Research Conference, Plasmonics and Nanophotonics (2024).
4. 'Empowering quantum devices with generative AI-topological latent diffusion model', DOE Quantum Science Center (QSC) All-Hands meeting (2024).
5. 'Empowering quantum 2.0 devices and approaches with machine learning', Optica Quantum 2.0 Conference and Exhibition (2022).

Other Experiences

1. Reviewer for ACS Photonics, Nanophotonics, PNAS, et al.
2. President of Purdue IEEE Photonics Society Student Chapter
3. Vice President of Purdue SPIE Student Chapter