Event Date
Event Date
Reconstructive spectrometers are getting traction for their high performance in a miniaturized form-factor. Although, the lack of bulky optics reduces the footprint of the reconstructive spectrometer, there still remains the trade-off between spectral resolution and spectral range. We present an innovative machine learning approach to circumvent this trade-off achieving 2.5× improved results than conventional approaches. The results are verified using an on-chip spectrometer that comprises of 16 unique silicon photodiodes spectrally engineered with photon-trapping nanostructures. The spectrometer achieves ~4 nm spectral resolution and ~100% peak accuracy while operating in the spectral range of 640 – 1100 nm making it practical for emerging scientific, communication, biomedical and consumer applications.
Presenter
Univ. of California, Davis (United States)
Dr. Ahasan Ahamed is a postdoctoral scholar in the University of California, Davis. He completed his doctoral studies in electrical and computer engineering from the University of California, Davis in 2025 focusing on reconstructive spectrometers using silicon photodiodes. He completed his M.Sc. degree from UC Davis in Electrical and Computer Engineering and his B.Sc. degree from Bangladesh University of Engineering and Technology (BUET) in Electrical and Electronics Engineering. His current research interests include optoelectronic sensors, silicon photonics, hyperspectral imaging, hardware security and photonic true random number generators. He received prestigious Smita Bakshi teaching award and Graduate Student Leadership award for his contribution to the academic and social community. He is actively participating outreach activities to promote STEM education through CITRIS-INSPIRE and GREAT program.