-
Weakly-Supervised Semantic Segmentation of Circular-Scan, Synthetic-Aperture-Sonar Imagery
Authors:
Isaac J. Sledge,
Dominic M. Byrne,
Jonathan L. King,
Steven H. Ostertag,
Denton L. Woods,
James L. Prater,
Jermaine L. Kennedy,
Timothy M. Marston,
Jose C. Principe
Abstract:
We propose a weakly-supervised framework for the semantic segmentation of circular-scan synthetic-aperture-sonar (CSAS) imagery. The first part of our framework is trained in a supervised manner, on image-level labels, to uncover a set of semi-sparse, spatially-discriminative regions in each image. The classification uncertainty of each region is then evaluated. Those areas with the lowest uncerta…
▽ More
We propose a weakly-supervised framework for the semantic segmentation of circular-scan synthetic-aperture-sonar (CSAS) imagery. The first part of our framework is trained in a supervised manner, on image-level labels, to uncover a set of semi-sparse, spatially-discriminative regions in each image. The classification uncertainty of each region is then evaluated. Those areas with the lowest uncertainties are then chosen to be weakly labeled segmentation seeds, at the pixel level, for the second part of the framework. Each of the seed extents are progressively resized according to an unsupervised, information-theoretic loss with structured-prediction regularizers. This resha** process uses multi-scale, adaptively-weighted features to delineate class-specific transitions in local image content. Content-addressable memories are inserted at various parts of our framework so that it can leverage features from previously seen images to improve segmentation performance for related images.
We evaluate our weakly-supervised framework using real-world CSAS imagery that contains over ten seafloor classes and ten target classes. We show that our framework performs comparably to nine fully-supervised deep networks. Our framework also outperforms eleven of the best weakly-supervised deep networks. We achieve state-of-the-art performance when pre-training on natural imagery. The average absolute performance gap to the next-best weakly-supervised network is well over ten percent for both natural imagery and sonar imagery. This gap is found to be statistically significant.
△ Less
Submitted 20 January, 2024;
originally announced January 2024.
-
Target Detection and Segmentation in Circular-Scan Synthetic-Aperture-Sonar Images using Semi-Supervised Convolutional Encoder-Decoders
Authors:
Isaac J. Sledge,
Matthew S. Emigh,
Jonathan L. King,
Denton L. Woods,
J. Tory Cobb,
Jose C. Principe
Abstract:
We propose a framework for saliency-based, multi-target detection and segmentation of circular-scan, synthetic-aperture-sonar (CSAS) imagery. Our framework relies on a multi-branch, convolutional encoder-decoder network (MB-CEDN). The encoder portion of the MB-CEDN extracts visual contrast features from CSAS images. These features are fed into dual decoders that perform pixel-level segmentation to…
▽ More
We propose a framework for saliency-based, multi-target detection and segmentation of circular-scan, synthetic-aperture-sonar (CSAS) imagery. Our framework relies on a multi-branch, convolutional encoder-decoder network (MB-CEDN). The encoder portion of the MB-CEDN extracts visual contrast features from CSAS images. These features are fed into dual decoders that perform pixel-level segmentation to mask targets. Each decoder provides different perspectives as to what constitutes a salient target. These opinions are aggregated and cascaded into a deep-parsing network to refine the segmentation.
We evaluate our framework using real-world CSAS imagery consisting of five broad target classes. We compare against existing approaches from the computer-vision literature. We show that our framework outperforms supervised, deep-saliency networks designed for natural imagery. It greatly outperforms unsupervised saliency approaches developed for natural imagery. This illustrates that natural-image-based models may need to be altered to be effective for this imaging-sonar modality.
△ Less
Submitted 17 February, 2022; v1 submitted 10 January, 2021;
originally announced January 2021.
-
Correcting thermal-emission-induced detector saturation in infrared reflection or transmission spectroscopy
Authors:
C. Yao,
H. Mei,
Y. Xiao,
A. Shahsafi,
W. Derdeyn,
J. L. King,
C. Wan,
R. O. Scarlat,
M. H. Anderson,
M. A. Kats
Abstract:
We found that temperature-dependent infrared spectroscopy measurements (i.e., reflectance or transmittance) using a Fourier-transform spectrometer can have substantial errors, especially for elevated sample temperatures and collection using an objective lens (e.g., using an infrared microscope). These errors arise as a result of partial detector saturation due to thermal emission from the measured…
▽ More
We found that temperature-dependent infrared spectroscopy measurements (i.e., reflectance or transmittance) using a Fourier-transform spectrometer can have substantial errors, especially for elevated sample temperatures and collection using an objective lens (e.g., using an infrared microscope). These errors arise as a result of partial detector saturation due to thermal emission from the measured sample reaching the detector, resulting in nonphysical apparent reduction of reflectance or transmittance with increasing sample temperature. Here, we demonstrate that these temperature-dependent errors can be corrected by implementing several levels of optical attenuation that enable "convergence testing" of the measured reflectance or transmittance as the thermal-emission signal is reduced, or by applying correction factors that can be inferred by looking at the spectral regions where the sample is not expected to have a substantial temperature dependence.
△ Less
Submitted 25 March, 2021; v1 submitted 29 December, 2020;
originally announced December 2020.
-
Impact of corrosion on the emissivity of advanced-reactor structural alloys
Authors:
Jonathan L. King,
Hang** Jo,
Alireza Shahsafi,
Kyle Blomstrand,
Kumar Sridharan,
Mikhail A. Kats
Abstract:
Under standard operating conditions, the emissivity of structural alloys used for various components of nuclear reactors may evolve, affecting the heat transfer of the systems. In this study, mid-infrared emissivities of several reactor structural alloys were measured before and after exposure to environments relevant to next-generation reactors. We evaluated nickel-based alloys Haynes 230 and Inc…
▽ More
Under standard operating conditions, the emissivity of structural alloys used for various components of nuclear reactors may evolve, affecting the heat transfer of the systems. In this study, mid-infrared emissivities of several reactor structural alloys were measured before and after exposure to environments relevant to next-generation reactors. We evaluated nickel-based alloys Haynes 230 and Inconel 617 exposed to helium gas at 1000 $^{\circ}$C, nickel-based Hastelloy N and iron-based 316 stainless steel exposed to molten salts at 750-850 $^{\circ}$C, 316 stainless steel exposed to liquid sodium at 650 $^{\circ}$C, and 316 stainless steel and Haynes 230 exposed to supercritical CO2 at 650 $^{\circ}$C. Emissivity was measured via emissive and reflective techniques using a Fourier transform infrared (FTIR) spectrometer. Large increases in emissivity are observed for alloys exposed to oxidizing environments, while only minor differences were observed in other exposure conditions.
△ Less
Submitted 11 May, 2018;
originally announced May 2018.
-
Atomic swelling upon compression
Authors:
V. K. Dolmatov,
J. L. King
Abstract:
The hydrogen atom under the pressure of a spherical penetrable confinement potential of a decreasing radius $r_{0}$ is explored, as a case study. A novel counter-intuitive effect of atomic swelling rather than shrinking with decreasing $r_{0}$ is unraveled, when $r_{0}$ reaches, and remains smaller than, a certain critical value. Upon swelling, the size of the atom is shown to increase by an order…
▽ More
The hydrogen atom under the pressure of a spherical penetrable confinement potential of a decreasing radius $r_{0}$ is explored, as a case study. A novel counter-intuitive effect of atomic swelling rather than shrinking with decreasing $r_{0}$ is unraveled, when $r_{0}$ reaches, and remains smaller than, a certain critical value. Upon swelling, the size of the atom is shown to increase by an order of magnitude, or more, compared to the size of the free atom. Examples of changes of photoabsorption properties of confined hydrogen atom upon its swelling are uncovered and demonstrated.
△ Less
Submitted 18 August, 2012;
originally announced August 2012.
-
Diffuse versus square-well confining potentials in modelling $A$@C$_{60}$ atoms
Authors:
V. K. Dolmatov,
J. L. King,
J. C. Oglesby
Abstract:
Attention: this version-$2$ of the manuscript differs from its previously uploaded version-$1$ (arXiv:1112.6158v1) and subsequently published in 2012 J. Phys. B \textbf{45} 105102 only by a removed typo in Eq.(2) of version-$1$; there was the erroneous factor "2" in both terms in the right-hand-side of the Eq.(2) of version-$1$. Now that the typo is removed, Eq.(2) is correct.
A perceived advant…
▽ More
Attention: this version-$2$ of the manuscript differs from its previously uploaded version-$1$ (arXiv:1112.6158v1) and subsequently published in 2012 J. Phys. B \textbf{45} 105102 only by a removed typo in Eq.(2) of version-$1$; there was the erroneous factor "2" in both terms in the right-hand-side of the Eq.(2) of version-$1$. Now that the typo is removed, Eq.(2) is correct.
A perceived advantage for the replacement of a discontinuous square-well pseudo-potential, which is often used by various researchers as an approximation to the actual C$_{60}$ cage potential in calculations of endohedral atoms $A$@C$_{60}$, by a more realistic diffuse potential is explored. The photoionization of endohedral H@C$_{60}$ and Xe@C$_{60}$ is chosen as the case study. The diffuse potential is modelled by a combination of two Woods-Saxon potentials. It is demonstrated that photoionization spectra of $A$@C$_{60}$ atoms are largely insensitive to the degree $η$ of diffuseness of the potential borders, in a reasonably broad range of $η$'s. Alternatively, these spectra are found to be insensitive to discontinuity of the square-well potential either. Both potentials result in practically identical calculated spectra. New numerical values for the set of square-well parameters, which lead to a better agreement between experimental and theoretical data for $A$@C$_{60}$ spectra, are recommended for future studies.
△ Less
Submitted 20 December, 2014; v1 submitted 28 December, 2011;
originally announced December 2011.
-
A change-of-coordinates from Geometry to Algebra, applied to Brick Tilings
Authors:
Jonathan L. King
Abstract:
A proof is sketched of the Polynomial Conjecture of the author (circulated as preprint "Brick Tiling and Monotone Boolean Functions", available at the http://www.math.ufl.edu/~squash/tilingstuff.html url) which says that the family of minimal tilable-boxes grows polynomially with dimension. An important ingredient of the argument is translating the problem from its finite-dimensional geometric f…
▽ More
A proof is sketched of the Polynomial Conjecture of the author (circulated as preprint "Brick Tiling and Monotone Boolean Functions", available at the http://www.math.ufl.edu/~squash/tilingstuff.html url) which says that the family of minimal tilable-boxes grows polynomially with dimension. An important ingredient of the argument is translating the problem from its finite-dimensional geometric framework to the algebraic setting of an infinite-dimensional lattice.
△ Less
Submitted 29 September, 1998;
originally announced September 1998.
