Thermal-Face-Contrastive GAN (TFC-GAN): A Framework for Visible-to-Thermal Face Generation
- Award ID(s):
- 1948399
- PAR ID:
- 10404941
- Date Published:
- Journal Name:
- ICML 2022 The 1st Workshop on Healthcare AI and COVID-19
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
Agaian, Sos S.; Jassim, Sabah A. (Ed.)Face recognition technologies have been in high demand in the past few decades due to the increase in human-computer interactions. It is also one of the essential components in interpreting human emotions, intentions, facial expressions for smart environments. This non-intrusive biometric authentication system relies on identifying unique facial features and pairing alike structures for identification and recognition. Application areas of facial recognition systems include homeland and border security, identification for law enforcement, access control to secure networks, authentication for online banking and video surveillance. While it is easy for humans to recognize faces under varying illumination conditions, it is still a challenging task in computer vision. Non-uniform illumination and uncontrolled operating environments can impair the performance of visual-spectrum based recognition systems. To address these difficulties, a novel Anisotropic Gradient Facial Recognition (AGFR) system that is capable of autonomous thermal infrared to visible face recognition is proposed. The main contribution of this paper includes a framework for thermal/fused-thermal-visible to visible face recognition system and a novel human-visual-system inspired thermal-visible image fusion technique. Extensive computer simulations using CARL, IRIS, AT&T, Yale and Yale-B databases demonstrate the efficiency, accuracy, and robustness of the AGFR system. Keywords: Infrared thermal to visible facial recognition, anisotropic gradient, visible-to-visible face recognition, nonuniform illumination face recognition, thermal and visible face fusion methodmore » « less
-
GaN samples were implanted with Be and annealed in different conditions in order to activate the shallow BeGaacceptor. Low-temperature photoluminescence spectra were studied to find BeGa-related defects in the implanted samples. A yellow band with a maximum at about 2.2 eV (the YLBeband) was observed in nearly all samples protected with an AlN cap during the annealing and in samples annealed under ultrahigh N2pressure. A green band with a maximum at 2.35 eV (the GL2 band), attributed to the nitrogen vacancy, was the dominant defect-related luminescence band in GaN samples annealed without a protective AlN layer. The ultraviolet luminescence (UVLBe) band with a maximum at 3.38 eV attributed to the shallow BeGaacceptor with the ionization energy of 0.113 eV appeared in implanted samples only after annealing at high temperatures and ultrahigh N2pressure. This is the first observation of the UVLBeband in Be-implanted GaN, indicating successful activation of the BeGaacceptor.more » « less
-
NiO is a promising alternative to p-GaN as a hole injection layer for normally-off lateral transistors or low on-resistance vertical heterojunction rectifiers. The valence band offsets of sputtered NiO on c-plane, vertical geometry homoepitaxial GaN structures were measured by x-ray photoelectron spectroscopy as a function of annealing temperatures to 600 °C. This allowed determination of the band alignment from the measured bandgap of NiO. This alignment was type II, staggered gap for both as-deposited and annealed samples. For as-deposited heterojunction, ΔEV = 2.89 eV and ΔEC = −2.39 eV, while for all the annealed samples, ΔEVvalues were in the range of 3.2–3.4 eV and ΔECvalues were in the range of −(2.87–3.05) eV. The bandgap of NiO was reduced from 3.90 eV as-deposited to 3.72 eV after 600 °C annealing, which accounts for much of the absolute change in ΔEV − ΔEC. At least some of the spread in reported band offsets for the NiO/GaN system may arise from differences in their thermal history.more » « less
An official website of the United States government

