MWIRGAN: Unsupervised Visible-to-MWIR Image Translation with Generative Adversarial Network

Uddin, M.S.; Kwan, Chiman; Li, Jiang

doi:10.3390/electronics12041039

Cited by 4 publications

(3 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There are examples of EO-IR devices that utilize visible-MWIR exclusively. 15 So this material could provide an alternative that is easier to shape and quicker to manufacture than existing visible-MWIR materials, like ALON. To characterize the doped glasses, different concentrations of silicon nitride dopant were introduced in GLSO, with the optical and mechanical properties tested and compared to the GLSO without dopants.…”

Section: Development New Materials For Opticsmentioning

confidence: 99%

Expanding the transmission window of visible-MWIR chalcogenide glasses by silicon nitride doping

Archer

Craig

et al. 2023

Window and Dome Technologies and Materials XVII

View full text Add to dashboard Cite

Gallium Lanthanum sulphur-oxide (GLSO) glass is an excellent candidate for a window/dome material ought to its wideband transmission window from visible to MWIR. The suitable optical transmission from 0.45-8 microns is supplemented by its superior thermal and mechanical properties to contemporary materials, such as Cleartran™ zinc sulphide. In this work, the properties of GLSO were enhanced by doping with small concentrations of silicon nitride (≤ 0.5 M%), which expanded the transmission window to encompass all the visible spectra. Nano-indentation demonstrated that the hardness and elastic modulus slightly improved. Overall, the improvements demonstrated here make this glass an even better solution when compared to the state-of-the-art for use in single-optic windows and domes.

show abstract

Section: Development New Materials For Opticsmentioning

confidence: 99%

Expanding the transmission window of visible-MWIR chalcogenide glasses by silicon nitride doping

Archer

Craig

et al. 2023

Window and Dome Technologies and Materials XVII

View full text Add to dashboard Cite

show abstract

“…The lack of open-source EO/IR datasets presents a strong motive to generate complementary synthetic data for machine learning. In several previous articles [10][11][12] this was attempted by training generative adversarial networks (GANs) to convert RGB data into IR data. Furthermore, the authors of Ref.…”

Section: Summary Of Deep Learning-based Vehicle Recognition Proceduresmentioning

confidence: 99%

Sensitivity analysis of ResNet-based automatic target recognition performance using MuSES-generated EO/IR synthetic imagery

Klein,

Young,

Taylor

et al. 2023

Synthetic Data for Artificial Intelligence and Machine Learning: Tools, Techniques, and Applications

View full text Add to dashboard Cite

Machine learning algorithms have demonstrated state-of-the-art automated target recognition performance but require a large training set. In the case of electro-optical infrared (EO/IR) remote sensing, acquiring sufficient measured imagery can be difficult, but EO/IR scene simulation is a possible alternative. CoTherm, a co-simulation tool which operates MuSES in an automated fashion, is used to manipulate relevant target, background and sensor inputs to generate a library of radiance images. Various options affecting simulation run-time and output fidelity are considered and the trade-off between accuracy and compute time requirements is quantified using a measured imagery benchmark and ResNets for classification.

show abstract

“…And data collection and annotation has been a long-standing problem, which is both expensive and time consuming. There are various related works to overcome this problem by using GAN-based methods [2] or training encoder-decoder networks [3], which utilize DSIAC database and generate additional infrared images.…”

Section: Introductionmentioning

confidence: 99%

Analyzing DSIAC ATR algorithm development database utilizing transfer learning

Özertem

2024

SPIE Future Sensing Technologies 2024

View full text Add to dashboard Cite

For military applications, recognizing the targets with a good accuracy is a vital skill. In the literature there are many machine learning-based works on target recognition in visible spectrum, since there are massive RGB datasets. However, it is very important to have the same capability in infrared spectrum for military applications. Because of that reason DSIAC database, which has both visible and infrared images of the same targets is introduced first. Then a straightforward and efficient transfer learning-based ATR algorithm is proposed. Each step of the transfer learning process is explained in detail. The proposed transfer learning algorithm is tested with many challenging scenarios of DSIAC database. We extract valuable results how the ATR performance depends on range, wavelength and time changes. We also test ATR capability of our proposed model against extensive data. At the end we achieve very satisfactory accuracy scores thanks to the power of transfer learning.

show abstract

MWIRGAN: Unsupervised Visible-to-MWIR Image Translation with Generative Adversarial Network

Cited by 4 publications

References 31 publications

Expanding the transmission window of visible-MWIR chalcogenide glasses by silicon nitride doping

Expanding the transmission window of visible-MWIR chalcogenide glasses by silicon nitride doping

Sensitivity analysis of ResNet-based automatic target recognition performance using MuSES-generated EO/IR synthetic imagery

Analyzing DSIAC ATR algorithm development database utilizing transfer learning

Contact Info

Product

Resources

About