2026
Anushka Kumari; Anushree Khisti; Abhinav Choube; Devansh Satra; Srivatsa Murali; Anshuman Kumar
Multimodal Optical Feature Extraction with a Free-Space Photonic Extreme Learning Machine Journal Article
In: arXiv preprint, vol. 2605.29043 , 2026.
@article{pelm,
title = {Multimodal Optical Feature Extraction with a Free-Space Photonic Extreme Learning Machine},
author = {Anushka Kumari and Anushree Khisti and Abhinav Choube and Devansh Satra and Srivatsa Murali and Anshuman Kumar},
url = {https://doi.org/10.48550/arXiv.2605.29043},
doi = {10.48550/arXiv.2605.29043},
year = {2026},
date = {2026-05-29},
urldate = {2026-05-29},
journal = {arXiv preprint},
volume = {2605.29043 },
abstract = {Photonic extreme learning machines (PELMs) replace a digitally trained hidden layer by a fixed optical transformation, allowing a high dimensional feature map to be generated by physical propagation while only the final readout is learned. Existing free-space PELM demonstrations have established this principle for image and tabular benchmarks, but a unified multimodal optical feature extractor spanning structurally different data types has remained largely undeveloped. Here we demonstrate a single free-space PELM platform for image, audio derived, binary tabular, and regression tasks using phase only SLM encoding, Fourier like free space propagation, and camera intensity detection.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Photonic extreme learning machines (PELMs) replace a digitally trained hidden layer by a fixed optical transformation, allowing a high dimensional feature map to be generated by physical propagation while only the final readout is learned. Existing free-space PELM demonstrations have established this principle for image and tabular benchmarks, but a unified multimodal optical feature extractor spanning structurally different data types has remained largely undeveloped. Here we demonstrate a single free-space PELM platform for image, audio derived, binary tabular, and regression tasks using phase only SLM encoding, Fourier like free space propagation, and camera intensity detection.
Vineetha Joy; Jamshed Palai; Satwik Sahoo; Anshuman Kumar; Amit Sethi; Hema Singh
Inverse Design of Metasurface based Absorbers using Physics Guided Conditional Diffusion Models Journal Article
In: arXiv preprint, vol. arXiv:2605.19611, 2026.
@article{arXiv:2605.19611,
title = {Inverse Design of Metasurface based Absorbers using Physics Guided Conditional Diffusion Models},
author = {Vineetha Joy and Jamshed Palai and Satwik Sahoo and Anshuman Kumar and Amit Sethi and Hema Singh},
url = {https://doi.org/10.48550/arXiv.2605.19611},
doi = {10.48550/arXiv.2605.19611},
year = {2026},
date = {2026-05-19},
journal = {arXiv preprint},
volume = {arXiv:2605.19611},
abstract = {Inverse design of metasurfaces for specific electromagnetic responses requires generating geometries that satisfy stringent spectral constraints while maintaining manufacturability. Conventional design methodologies rely on iterative optimization routines using full wave simulations, which become extremely time consuming and computationally intensive for large design spaces. In addition, commonly employed generative approaches often exhibit limited conditional fidelity and the generated designs often contain fine or irregular features that are impractical to fabricate. In this regard, we propose a physics guided condition quality enhanced diffusion framework for the inverse design of metasurface based absorbers. Here, the conditioning information consisting of target reflection characteristics is integrated into the model using feature wise linear modulation (FiLM). Furthermore, to enforce adherence to target spectra, a pre trained surrogate EM simulator is embedded into the framework introducing physics aware regularization through spectrum level loss functions. The efficiency of the proposed model is demonstrated by generating practically realizable metasurfaces for different types of reflection characteristics in the frequency range of 2 to 18 GHz. The proposed framework achieves an average spectral mean squared error of 0.0006 and band alignment accuracy of 0.958 between the target spectra and the spectra produced by the generated designs, demonstrating high conditional accuracy. In addition, the model generates multiple geometries for the same condition, thereby providing diverse design alternatives to the engineer. The proposed model produces the suitable design in approximately 30 seconds, whereas the conventional approach can take several months under comparable computational resources. The efficiency of the model is also established via experimental measurements.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Inverse design of metasurfaces for specific electromagnetic responses requires generating geometries that satisfy stringent spectral constraints while maintaining manufacturability. Conventional design methodologies rely on iterative optimization routines using full wave simulations, which become extremely time consuming and computationally intensive for large design spaces. In addition, commonly employed generative approaches often exhibit limited conditional fidelity and the generated designs often contain fine or irregular features that are impractical to fabricate. In this regard, we propose a physics guided condition quality enhanced diffusion framework for the inverse design of metasurface based absorbers. Here, the conditioning information consisting of target reflection characteristics is integrated into the model using feature wise linear modulation (FiLM). Furthermore, to enforce adherence to target spectra, a pre trained surrogate EM simulator is embedded into the framework introducing physics aware regularization through spectrum level loss functions. The efficiency of the proposed model is demonstrated by generating practically realizable metasurfaces for different types of reflection characteristics in the frequency range of 2 to 18 GHz. The proposed framework achieves an average spectral mean squared error of 0.0006 and band alignment accuracy of 0.958 between the target spectra and the spectra produced by the generated designs, demonstrating high conditional accuracy. In addition, the model generates multiple geometries for the same condition, thereby providing diverse design alternatives to the engineer. The proposed model produces the suitable design in approximately 30 seconds, whereas the conventional approach can take several months under comparable computational resources. The efficiency of the model is also established via experimental measurements.
Devansh Satra; Abhishek Kumar; Anshuman Kumar
Inverse Design of the Topology–Bandwidth Tradeoff in Valley Photonic Crystals Journal Article
In: arXiv preprint, vol. arXiv:2601.22958 , 2026.
@article{arXiv:2601.22958,
title = {Inverse Design of the Topology–Bandwidth Tradeoff in Valley Photonic Crystals},
author = {Devansh Satra and Abhishek Kumar and Anshuman Kumar},
url = {https://doi.org/10.48550/arXiv.2601.22958},
doi = {10.48550/arXiv.2601.22958},
year = {2026},
date = {2026-01-30},
urldate = {2026-01-30},
journal = {arXiv preprint},
volume = {arXiv:2601.22958 },
abstract = {Integrated on-chip photonics increasingly relies on wave propagation that remains stable in the presence of fabrication imperfections, tight bends, and dense routing. Valley photonic crystals (VPCs) offer an attractive path: by opening a gap at the Dirac points of a hexagonal lattice, one can engineer guided modes confined to domain walls that thread around corners with reduced backreflection. We develop a design framework that co-optimizes the photonic bulk band gap and valley Chern number using a modified particle-swarm optimization (PSO), while evaluating the photonic band structure via plane-wave expansion and the topological characteristics using a gauge-invariant lattice discretization to compute the Berry-curvature. The optimized structures exhibit a clean valley-Hall gap with edge bands traversing the gap and high interface transmission in full-wave simulations. These results consolidate topology-aware geometry optimization for robust on-chip guiding.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Integrated on-chip photonics increasingly relies on wave propagation that remains stable in the presence of fabrication imperfections, tight bends, and dense routing. Valley photonic crystals (VPCs) offer an attractive path: by opening a gap at the Dirac points of a hexagonal lattice, one can engineer guided modes confined to domain walls that thread around corners with reduced backreflection. We develop a design framework that co-optimizes the photonic bulk band gap and valley Chern number using a modified particle-swarm optimization (PSO), while evaluating the photonic band structure via plane-wave expansion and the topological characteristics using a gauge-invariant lattice discretization to compute the Berry-curvature. The optimized structures exhibit a clean valley-Hall gap with edge bands traversing the gap and high interface transmission in full-wave simulations. These results consolidate topology-aware geometry optimization for robust on-chip guiding.
2025
Vineetha Joy; Aditya Anand; Nidhi; Anshuman Kumar; Amit Sethi; Hema Singh
A Surrogate Model for the Forward Design of Multi-layered Metasurface-based Radar Absorbing Structures Journal Article
In: arXiv preprint, vol. arXiv:2505.09251, 2025.
@article{arXiv:2505.09251,
title = {A Surrogate Model for the Forward Design of Multi-layered Metasurface-based Radar Absorbing Structures},
author = {Vineetha Joy and Aditya Anand and Nidhi and Anshuman Kumar and Amit Sethi and Hema Singh},
url = {https://doi.org/10.48550/arXiv.2505.09251},
doi = {10.48550/arXiv.2505.09251},
year = {2025},
date = {2025-05-14},
journal = {arXiv preprint},
volume = {arXiv:2505.09251},
abstract = {Metasurface-based radar absorbing structures (RAS) are highly preferred for applications like stealth technology, electromagnetic (EM) shielding, etc. due to their capability to achieve frequency selective absorption characteristics with minimal thickness and reduced weight penalty. However, the conventional approach for the EM design and optimization of these structures relies on forward simulations, using full wave simulation tools, to predict the electromagnetic (EM) response of candidate meta atoms. This process is computationally intensive, extremely time consuming and requires exploration of large design spaces. To overcome this challenge, we propose a surrogate model that significantly accelerates the prediction of EM responses of multi-layered metasurface-based RAS. A convolutional neural network (CNN) based architecture with Huber loss function has been employed to estimate the reflection characteristics of the RAS model. The proposed model achieved a cosine similarity of 99.9 percent and a mean square error of 0.001 within 1000 epochs of training. The efficiency of the model has been established via full wave simulations as well as experiment where it demonstrated significant reduction in computational time while maintaining high predictive accuracy.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Metasurface-based radar absorbing structures (RAS) are highly preferred for applications like stealth technology, electromagnetic (EM) shielding, etc. due to their capability to achieve frequency selective absorption characteristics with minimal thickness and reduced weight penalty. However, the conventional approach for the EM design and optimization of these structures relies on forward simulations, using full wave simulation tools, to predict the electromagnetic (EM) response of candidate meta atoms. This process is computationally intensive, extremely time consuming and requires exploration of large design spaces. To overcome this challenge, we propose a surrogate model that significantly accelerates the prediction of EM responses of multi-layered metasurface-based RAS. A convolutional neural network (CNN) based architecture with Huber loss function has been employed to estimate the reflection characteristics of the RAS model. The proposed model achieved a cosine similarity of 99.9 percent and a mean square error of 0.001 within 1000 epochs of training. The efficiency of the model has been established via full wave simulations as well as experiment where it demonstrated significant reduction in computational time while maintaining high predictive accuracy.
2024
Nihar Ranjan Sahoo; S S Jatin Prasath; Brijesh Kumar; Anshuman Kumar
Twisted hyperbolic van der Waals crystals for chip-scale full Stokes mid-infrared polarization detection Journal Article
In: Journal of Physics D: Applied Physics, vol. 57, no. 50, pp. 505104, 2024.
@article{stokes_polarimeter,
title = {Twisted hyperbolic van der Waals crystals for chip-scale full Stokes mid-infrared polarization detection},
author = {Nihar Ranjan Sahoo and S S Jatin Prasath and Brijesh Kumar and Anshuman Kumar},
url = {https://iopscience.iop.org/article/10.1088/1361-6463/ad7a85},
doi = {10.1088/1361-6463/ad7a85},
year = {2024},
date = {2024-09-25},
urldate = {2024-09-25},
journal = {Journal of Physics D: Applied Physics},
volume = {57},
number = {50},
pages = {505104},
abstract = {Investigating the polarization properties of light in the mid-infrared (mid-IR) spectrum is crucial for molecular sensing, biomedical diagnostics, and IR imaging system technologies. Traditional methods, limited by bulky size and complicated fabrication process, utilize large rotating optics for full Stokes polarization detection, impeding miniaturization and accuracy. Naturally occurring hyperbolic van der Waals (vdW) material based devices can address these challenges due to their lithography-free fabrication, ease of integration with chip-scale platforms and room-temperature operation. This study designs a chip-integrated polarimeter by performing multi-objective optimization for efficient exploration of the design parameter space. The spatial division measurement scheme used incorporates six precisely designed linear and circular polarization filters, achieving high extinction ratios exceeding 30 dB and transmittance surpassing 50
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Investigating the polarization properties of light in the mid-infrared (mid-IR) spectrum is crucial for molecular sensing, biomedical diagnostics, and IR imaging system technologies. Traditional methods, limited by bulky size and complicated fabrication process, utilize large rotating optics for full Stokes polarization detection, impeding miniaturization and accuracy. Naturally occurring hyperbolic van der Waals (vdW) material based devices can address these challenges due to their lithography-free fabrication, ease of integration with chip-scale platforms and room-temperature operation. This study designs a chip-integrated polarimeter by performing multi-objective optimization for efficient exploration of the design parameter space. The spatial division measurement scheme used incorporates six precisely designed linear and circular polarization filters, achieving high extinction ratios exceeding 30 dB and transmittance surpassing 50
title = {Deep learning-based variational autoencoder for classification of quantum and classical states of light},
author = {Mahesh Bhupati and Abhishek Mall and Anshuman Kumar and Pankaj K Jha},
url = {https://doi.org/10.48550/arXiv.2405.05243},
doi = {10.48550/arXiv.2405.05243},
year = {2024},
date = {2024-05-08},
urldate = {2024-05-08},
journal = {arXiv preprint},
volume = {arXiv:2405.05243},
abstract = {Advancements in optical quantum technologies have been enabled by the generation, manipulation, and characterization of light, with identification based on its photon statistics. However, characterizing light and its sources through single photon measurements often requires efficient detectors and longer measurement times to obtain high-quality photon statistics. Here we introduce a deep learning-based variational autoencoder (VAE) method for classifying single photon added coherent state (SPACS), single photon added thermal state (SPACS), mixed states between coherent/SPACS and thermal/SPATS of light. Our semisupervised learning-based VAE efficiently maps the photon statistics features of light to a lower dimension, enabling quasi-instantaneous classification with low average photon counts. The proposed VAE method is robust and maintains classification accuracy in the presence of losses inherent in an experiment, such as finite collection efficiency, non-unity quantum efficiency, finite number of detectors, etc. Additionally, leveraging the transfer learning capabilities of VAE enables successful classification of data of any quality using a single trained model. We envision that such a deep learning methodology will enable better classification of quantum light and light sources even in the presence of poor detection quality.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
author = {Mahesh Bhupati and Abhishek Mall and Anshuman Kumar and Pankaj K Jha},
url = {https://doi.org/10.48550/arXiv.2405.05243},
doi = {10.48550/arXiv.2405.05243},
year = {2024},
date = {2024-05-08},
urldate = {2024-05-08},
journal = {arXiv preprint},
volume = {arXiv:2405.05243},
abstract = {Advancements in optical quantum technologies have been enabled by the generation, manipulation, and characterization of light, with identification based on its photon statistics. However, characterizing light and its sources through single photon measurements often requires efficient detectors and longer measurement times to obtain high-quality photon statistics. Here we introduce a deep learning-based variational autoencoder (VAE) method for classifying single photon added coherent state (SPACS), single photon added thermal state (SPACS), mixed states between coherent/SPACS and thermal/SPATS of light. Our semisupervised learning-based VAE efficiently maps the photon statistics features of light to a lower dimension, enabling quasi-instantaneous classification with low average photon counts. The proposed VAE method is robust and maintains classification accuracy in the presence of losses inherent in an experiment, such as finite collection efficiency, non-unity quantum efficiency, finite number of detectors, etc. Additionally, leveraging the transfer learning capabilities of VAE enables successful classification of data of any quality using a single trained model. We envision that such a deep learning methodology will enable better classification of quantum light and light sources even in the presence of poor detection quality.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Advancements in optical quantum technologies have been enabled by the generation, manipulation, and characterization of light, with identification based on its photon statistics. However, characterizing light and its sources through single photon measurements often requires efficient detectors and longer measurement times to obtain high-quality photon statistics. Here we introduce a deep learning-based variational autoencoder (VAE) method for classifying single photon added coherent state (SPACS), single photon added thermal state (SPACS), mixed states between coherent/SPACS and thermal/SPATS of light. Our semisupervised learning-based VAE efficiently maps the photon statistics features of light to a lower dimension, enabling quasi-instantaneous classification with low average photon counts. The proposed VAE method is robust and maintains classification accuracy in the presence of losses inherent in an experiment, such as finite collection efficiency, non-unity quantum efficiency, finite number of detectors, etc. Additionally, leveraging the transfer learning capabilities of VAE enables successful classification of data of any quality using a single trained model. We envision that such a deep learning methodology will enable better classification of quantum light and light sources even in the presence of poor detection quality.
2020
Abhishek Mall, Abhijeet Patil, Amit Sethi, Anshuman Kumar
A Cyclical Deep Learning Based Framework For Simultaneous Inverse and Forward design of Nanophotonic Metasurfaces Journal Article
In: Scientific Reports, vol. 10, no. 1, pp. 19427, 2020.
@article{arXiv:2005.12796,
title = {A Cyclical Deep Learning Based Framework For Simultaneous Inverse and Forward design of Nanophotonic Metasurfaces},
author = {Abhishek Mall, Abhijeet Patil, Amit Sethi, Anshuman Kumar},
url = {https://www.nature.com/articles/s41598-020-76400-y},
doi = {10.1038/s41598-020-76400-y},
year = {2020},
date = {2020-11-10},
urldate = {2020-11-10},
journal = {Scientific Reports},
volume = {10},
number = {1},
pages = {19427},
abstract = {The conventional approach to nanophotonic metasurface design and optimization for a targeted electromagnetic response involves exploring large geometry and material spaces. This is a highly iterative process based on trial and error, which is computationally costly and time consuming. Moreover, the non-uniqueness of structural designs and high non-linearity between electromagnetic response and design makes this problem challenging. To model this unintuitive relationship between electromagnetic response and metasurface structural design as a probability distribution in the design space, we introduce a framework for inverse design of nanophotonic metasurfaces based on cyclical deep learning (DL). The proposed framework performs inverse design and optimization mechanism for the generation of meta-atoms and meta-molecules as metasurface units based on DL models and genetic algorithm. The framework includes consecutive DL models that emulate both numerical electromagnetic simulation and iterative processes of optimization, and generate optimized structural designs while simultaneously performing forward and inverse design tasks. A selection and evaluation of generated structural designs is performed by the genetic algorithm to construct a desired optical response and design space that mimics real world responses. Importantly, our cyclical generation framework also explores the space of new metasurface topologies. As an example application of the utility of our proposed architecture, we demonstrate the inverse design of gap-plasmon based half-wave plate metasurface for user-defined optical response. Our proposed technique can be easily generalized for designing nanophtonic metasurfaces for a wide range of targeted optical response.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The conventional approach to nanophotonic metasurface design and optimization for a targeted electromagnetic response involves exploring large geometry and material spaces. This is a highly iterative process based on trial and error, which is computationally costly and time consuming. Moreover, the non-uniqueness of structural designs and high non-linearity between electromagnetic response and design makes this problem challenging. To model this unintuitive relationship between electromagnetic response and metasurface structural design as a probability distribution in the design space, we introduce a framework for inverse design of nanophotonic metasurfaces based on cyclical deep learning (DL). The proposed framework performs inverse design and optimization mechanism for the generation of meta-atoms and meta-molecules as metasurface units based on DL models and genetic algorithm. The framework includes consecutive DL models that emulate both numerical electromagnetic simulation and iterative processes of optimization, and generate optimized structural designs while simultaneously performing forward and inverse design tasks. A selection and evaluation of generated structural designs is performed by the genetic algorithm to construct a desired optical response and design space that mimics real world responses. Importantly, our cyclical generation framework also explores the space of new metasurface topologies. As an example application of the utility of our proposed architecture, we demonstrate the inverse design of gap-plasmon based half-wave plate metasurface for user-defined optical response. Our proposed technique can be easily generalized for designing nanophtonic metasurfaces for a wide range of targeted optical response.
Abhishek Mall, Abhijeet Patil, Dipesh Tamboli, Amit Sethi, Anshuman Kumar
Fast Design of Plasmonic Metasurfaces Enabled by Deep Learning Journal Article
In: Journal of Physics D: Applied Physics, vol. 53, no. 49, pp. 49LT01, 2020.
@article{arXiv:2003.12402,
title = {Fast Design of Plasmonic Metasurfaces Enabled by Deep Learning},
author = {Abhishek Mall, Abhijeet Patil, Dipesh Tamboli, Amit Sethi, Anshuman Kumar},
url = {https://doi.org/10.1088/1361-6463/abb33c},
doi = {10.1088/1361-6463/abb33c},
year = {2020},
date = {2020-08-27},
urldate = {2020-08-27},
journal = {Journal of Physics D: Applied Physics},
volume = {53},
number = {49},
pages = {49LT01},
abstract = {Metasurfaces is an emerging field that enables the manipulation of light by an ultrathin structure composed of subwavelength antennae and fulfills an important requirement for miniaturized optical elements. Finding a new design for a metasurface or optimizing an existing design for a desired functionality is a computationally expensive and time consuming process as it is based on an iterative process of trial and error. We propose a deep learning (DL) architecture dubbed bidirectional autoencoder for nanophotonic metasurface design via a template search methodology. In contrast with the earlier approaches based on DL, our methodology addresses optimization in the space of multiple metasurface topologies instead of just one, in order to tackle the one to many mapping problem of inverse design. We demonstrate the creation of a Geometry and Parameter Space Library (GPSL) of metasurface designs with their corresponding optical response using our DL model. This GPSL acts as a universal design and response space for the optimization. As an example application, we use our methodology to design a multi-band gap-plasmon based half-wave plate metasurface. Through this example, we demonstrate the power of our technique in addressing the nonuniqueness problem of common inverse design approaches by showing that our network converges aptly to multiple metasurface topologies for the desired optical response. Our proposed technique would enable fast and accurate design and optimization of various kinds of metasurfaces with different functionalities.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Metasurfaces is an emerging field that enables the manipulation of light by an ultrathin structure composed of subwavelength antennae and fulfills an important requirement for miniaturized optical elements. Finding a new design for a metasurface or optimizing an existing design for a desired functionality is a computationally expensive and time consuming process as it is based on an iterative process of trial and error. We propose a deep learning (DL) architecture dubbed bidirectional autoencoder for nanophotonic metasurface design via a template search methodology. In contrast with the earlier approaches based on DL, our methodology addresses optimization in the space of multiple metasurface topologies instead of just one, in order to tackle the one to many mapping problem of inverse design. We demonstrate the creation of a Geometry and Parameter Space Library (GPSL) of metasurface designs with their corresponding optical response using our DL model. This GPSL acts as a universal design and response space for the optimization. As an example application, we use our methodology to design a multi-band gap-plasmon based half-wave plate metasurface. Through this example, we demonstrate the power of our technique in addressing the nonuniqueness problem of common inverse design approaches by showing that our network converges aptly to multiple metasurface topologies for the desired optical response. Our proposed technique would enable fast and accurate design and optimization of various kinds of metasurfaces with different functionalities.