ECE Research Seminar – Friday August 30 12-1 pm
August 26, 2024
Please join us for our August 2024 ECE Research Seminar
August 30, Friday, 12:00 – 1:00 pm, Simrall 104 and WebEx
https://msstate.webex.com/msstate/j.php?MTID=mc6aa170e551194fa8a5e605654429378
HRSpecNET: A Deep Learning-Based High-Resolution Radar Micro-Doppler
Signature Reconstruction for Improved HAR Classification
Sabyasachi Biswas | sb3682@msstate.edu
Abstract: Micro-Doppler signatures are widely used for radar-based human activity recognition (HAR). Traditional methods for generating micro-Doppler signatures, such as the Short-Time Fourier Transform (STFT), suffer from limitations, such as the trade-off between time and frequency resolution, noise sensitivity, and parameter calibration. To address these limitations, we propose a novel deep learning-based approach to reconstruct high-resolution micro-Doppler signatures directly from 1D complex time-domain signal. Our deep learning architecture consists of an autoencoder block to improve signal-to-noise ratio (SNR), an STFT block to learn frequency transformations to generate pseudo spectrograms, and finally, a UNET block to reconstruct high-resolution spectrogram images. We evaluated our proposed architecture on both synthetic and real-world data. For synthetic data, we generated 1D complex time domain signals with multiple time-varying frequencies and evaluated and compared the ability of our network to generate high-resolution micro-Doppler signatures and perform in different SNR levels. For real-world data, a challenging radar-based American Sign Language (ASL) dataset consisting of 100 words was used to evaluate the classification performance achieved by the proposed approach. The results showed that the proposed approach outperforms the traditional STFT-based micro-Doppler signatures by 3.48% in accuracy. Both synthetic and experimental micro-Doppler signatures show that the proposed approach learns to reconstruct higher-resolution and sparser spectrograms.
Mr. Sabyasachi Biswas received his B.Sc. degree in Electrical and Electronic Engineering from Bangladesh University of Engineering and Technology (BUET), Dhaka, Bangladesh, in 2019. He is currently pursuing his Ph.D. degree in Electrical and Computer Engineering at Mississippi State University (MSU). He is a research assistant at the Information Processing and Sensing (IMPRESS) Laboratory. His research interests include radar signal processing, human activity recognition using radar, camera, and lidar, and developing machine learning algorithms for activity classification using raw radar signals. He was a machine learning intern at the High-Performance Computing Collaboratory in Summer 2022. He was also a motion sensor and signal processing Co-Op at Bose from January to June 2024, where he developed a data driven machine learning algorithm for Bose Spatial Audio. He is a student member of the IEEE and also a member of the IEEE Signal Processing Society. He was the winner and 2nd runner-up of the Graduate Research Symposium in Fall 2022 and 2023, respectively, held at MSU.
* For further information contact: Dr. Jenny Du | du@ece.msstate.edu | 5-2035
For WebEx access, scan the QR code
The Department of Electrical and Computer Engineering at Mississippi State University consists of 27 faculty members (including seven endowed professors), seven professional staff, and over 700 undergraduate and graduate students, with approximately 100 being at the Ph.D. level. With a research expenditure of over $14.24 million, the department houses the largest High Voltage Laboratory among North American universities.