Sihan Shao (邵思涵)

I am a M.Sc. student, currently working at the Millimetre Wave and THz Techniques Lab. at Aalto University, advised by Prof. Zachary Taylor and D.Sc. Aleksi Tamminen. My research interests includes computational imaging in optics and THz, computer vision and antenna design .

I am pursuing a M.Sc. in Electronics and Nanotechnology (major) and Machine Learning (minor) from Aalto University, Finland. Previously, I obtained my B.Eng. degree in Opto-Electronics Information Science and Engineering at Changchun University of Science and Technology University (CUST), China.

In my free time, I enjoy running, swimming, and sauna.

Email  /  CV  /  GitHub  /  LinkedIn

Research

I'm interested in computational imaging, computer vision, machine learning, and optimization.

Neural Dispersive Hologram for Computational Submillimeter-wave Imaging

Developed a multi-wavelength wave propagation simulation using PyTorch, co-optimized a sub-millimeter imaging system’s hardware and neural networks, introduced a surrogate gradient method for optimizing quantized holograms, and validated the system with real measurements.

Quantized THz Diffractive Optics Design via Automatic Differentiation

We explored the angular spectrum method (ASM) with automatic differentiation to optimize highly quantized diffractive optical elements (DOEs) for terahertz applications, facilitating precise control over radiation modulation while accounting for practical manufacturing limitations of 3D printer.

Harnessing Frequency Diversity for Improved Holographic Imaging Systems

The study examines how frequency diversity over time in a 340 GHz FMCW radar system with a dispersive hologram enhances target imaging and discrimination by capturing more detailed spatial information.

SAR Ship Detection from Complex Background Based on Dynamic Shrinkage Attention Mechanism

We developed a SAR ship detection method addressing challenges like speckle noise and complex backgrounds using a feature-level denoising approach called dynamic shrinkage attention.

Soft Thresholding Attention Network for Adaptive Feature Denoising in SAR Ship Detection

Proposed a feature-level denoising methodology for synthetic aperture radar ship detection, leveraging soft thresholding and attention mechanisms, which enhanced detection precision across several neural network models.

Lidar Sensor-Based Object Recognition Using Machine Learning

We developed a Lidar-based data collection system and applied filtering, clustering, PCA, and SVM techniques to improve the classification efficiency and accuracy of pedestrians and vehicles in point cloud data.

Projects

These include coursework, thesis projects and unpublished research work.

Design of Smartwatch Bluetooth Rim Antenna

With Ali Kourani. Designed a smartwatch rim antenna integrated into the device’s lateral faces using CST and AWR software, tested with a hand phantom and battery cover.

Design of Vivaldi Array Antenna

With Guo Wei and Zhenhan Wang. Designed and manufactured a 4-Way Wilkinson Power Divider and Vivaldi array antenna with CST.

A 1.5 GHz Doppler Radar

With Yifan Yu. Designed and manufactured 1.5 GHz Doppler Radar with filters, mixer, oscillaor and amplifier.

Design and source code from Jon Barron's website