BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Asia/Tokyo
X-LIC-LOCATION:Asia/Tokyo
BEGIN:STANDARD
TZOFFSETFROM:+0900
TZOFFSETTO:+0900
TZNAME:JST
DTSTART:18871231T000000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260817T171533Z
LOCATION:Hall B5 (2)\, B Block\, Level 5
DTSTART;TZID=Asia/Tokyo:20241204T152700
DTEND;TZID=Asia/Tokyo:20241204T154100
UID:siggraphasia_SIGGRAPH Asia 2024_sess119_papers_918@linklings.com
SUMMARY:Learned Multi-aperture Color-coded Optics for Snapshot Hyperspectr
 al Imaging
DESCRIPTION:Zheng Shi (Princeton University); Xiong Dun (Tongji University
 ); Haoyu Wei (University of Hong Kong); Siyu Dong, Zhanshan Wang, and Xinb
 in Cheng (Tongji University); Felix Heide (Princeton University); and Yifa
 n Peng (University of Hong Kong)\n\nLearned optics, which incorporate ligh
 tweight diffractive optics, coded-aperture modulation, and specialized ima
 ge-processing neural networks, have recently garnered attention in the fie
 ld of snapshot hyperspectral imaging (HSI). While conventional methods typ
 ically rely on a single lens element paired with an off-the-shelf color se
 nsor, these setups, despite their widespread availability, present inheren
 t limitations. First, the Bayer sensor's spectral response curves are not 
 optimized for HSI applications, limiting spectral fidelity of the reconstr
 uction. Second, single lens designs rely on a single diffractive optical e
 lement (DOE)  to simultaneously encode spectral information and maintain s
 patial resolution across all wavelengths, which constrains spectral encodi
 ng capabilities.\nThis work investigates a multi-channel lens array combin
 ed with aperture-wise color filters, all co-optimized alongside an image r
 econstruction network. This configuration enables independent spatial enco
 ding and spectral response for each channel, improving optical encoding ac
 ross both spatial and spectral dimensions. Specifically, we validate that 
 the method achieves over a 5dB improvement in PSNR for spectral reconstruc
 tion compared to existing single-diffractive lens and coded-aperture techn
 iques. Experimental validation further confirmed that the method is capabl
 e of recovering up to 31 spectral bands within the 429–700~nm range in div
 erse indoor and outdoor environments.\n\nRegistration Category: Full Acces
 s, Full Access Supporter\n\nLanguage Format: English Language\n\nSession C
 hair: Qi Sun (New York University, Google)\n\n
URL:https://asia.siggraph.org/2024/program/?id=papers_918&sess=sess119
END:VEVENT
END:VCALENDAR
