IEEE GRSS Taipei Chapter invited Akira Hirose to present a talk on Advanced Neural Information Processing in Interferometric Synthtic Aperture Radar (InSAR) and Polarimetric SAR (PolSAR)
14 and 15 June 2017
Archive: --- IEEE.tv - Live Event - GRSS Live

Talk 1
Date & Time: June 14 (Wed), 2017, 13:00-14:40
Place: B2 International Hall, Hung-Yu Building, national Taipei University of Technology (NTUT = Taipei Tech), Taiwan
Host: Prof. Scott Yang-Lang Chang (Taipei Tech, IEEE GRSS Taipei Chapter Chair), Dr. Bryan C. Y. Chu (Taipei Tech)
Title: ”Advanced Neural Adaptive Processing in Interferometric and Polarimetric Radar Imaging”

The lecture was held in International Hall, Hung-Yu Building, national Taipei University of Technology, Taipei. About 250 audience attended to discuss the adaptive processing merits and its origin of complex-valued in Interferometric synthetic aperture radar (InSAR). The audience was the faculty and students of Taipei University of Technology majoring Electrical Engineering and Computer Science. In the lecture, I focused on the use of complex-valued neural networks (CVNN) in adaptive filtering of interferograms in InSAR. The audience expressed strong interest in the origin of the merits of complex-valued neural networks, in particular in the generalization ability suitable for wave information such as InSAR observation, which was illustrated in estimation of fringes in interferograms. Questions were concerning various aspects of the neural adaptive processing including the use of GPGPU in the processing and applicability of deep learning.

This DLP talk was broadcasted as live streaming on IEEE.tv for the first time in GRSS events, and also archived for further public viewing. The IEEE.tv observation showed very high statistics in the number of pageviews, registration and continuous audience. This event was made possible by the great effort and good teamwork of Taipei Tech staff, volunteer students, IEEE.tv staff and GRSS planners. The speaker is deeply grateful to all the people.

Archive:--- IEEE.tv - Live Event - GRSS Live


Talk 2
Date & Time: June 15 (Thu), 2017, 10:00-11:00
Place: R304 Delta Hall, Dept. of Electrical Engineering, National Tsing Hua University (NTHU), Taiwan
Host: Prof. Chong-Yung Chi, Prof. Yao-Win Peter Hong, Assoc. Prof. Che Lin (NTHU)
Title: ”Advanced Neural Adaptive Processing in Interferometric and Polarimetric Radar Imaging”

The lecture was held in Dept. of Electrical Engineering, National Tsing Hua University, Taiwan. About 30 audience attended to discuss the adaptive processing merits and its origin of complex-valued in Interferometric synthetic aperture radar (InSAR). The audience was the faculty and students of Taipei University of Technology majoring Electrical Engineering and Computer Science. In the lecture, I focused on the use of complex-valued neural networks (CVNN) in adaptive filtering of interferograms in InSAR. The audience expressed strong interest in the origin of the merits of complex-valued neural networks, in particular in the generalization ability suitable for wave information such as InSAR observation, which was illustrated in estimation of fringes in interferograms. Questions included neural activation functions and software implementation. University staff and students enjoyed wonderful discussion.


Talk 3
Date & Time: June 15 (Thu), 2017, 14:30-15:30
Place: R701 Yanping Technology Building, National Taiwan Ocean University (NTOU), Taiwan
Host: Prof. Tiong Kwong-Kau, Prof. Lena Chang (NTOU, IEEE Taipei Chapter Deputy Chair)
Title: ”Advanced Neural Adaptive Processing in Interferometric and Polarimetric Radar Imaging”

The lecture was held in Yanping Technology Building, National Taiwan Ocean University, Taiwan. About 50 audience attended to discuss the adaptive processing merits and its origin of complex-valued in Interferometric synthetic aperture radar (InSAR). The audience was the faculty and students of Taipei University of Technology majoring Electrical Engineering and Computer Science. In the lecture, I focused on the use of complex-valued neural networks (CVNN) in adaptive filtering of interferograms in InSAR. The audience expressed strong interest in the origin of the merits of complex-valued neural networks, in particular in the generalization ability suitable for wave information such as InSAR observation, which was illustrated in estimation of fringes in interferograms. Questions included observation geometry and electronics as well as neural network operation. Polarimetric SAR (PolSAR) was also discussed in the questions. We confirmed the effectiveness of CVNNs and QNNs in remote sensing fields.


Archive: Talk starting at around 45min in the video archive --- IEEE.tv - Live Event - GRSS Live