James A. Grant-Jacob
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
James A. Grant-Jacob has a total of 61 co-authors. Here are the top co-authors by number of shared publications:
- Ben Mills
University of Southampton
41 shared publications - Matthew Praeger
University of Southampton
18 shared publications - Robert W. Eason
Optoelectronics Research Centre, University of Southampton, Southampton, SO17 1BJ, United Kingdom
16 shared publications - Michalis N. Zervas
University of Southampton
14 shared publications Ben Mills
12 shared publications- Daniel J. Heath
ORC, University of Southampton Faculty of Physical Sciences and Engineering, Building 46, University of Southampton, Southampton, Hampshire, SO17 1BJ, UNITED KINGDOM OF GREAT BRITAIN AND NORTHERN IRELAND
11 shared publications
Latest Publications
- 1 December 2026
- Published byĀ Institute of Electrical and Electronics Engineers (IEEE)Ā inĀ IEEE Photonics Technology Letters
- Vol.Ā 38 Ā (23) , 1929-1932
Single-Image Inference of Power-Dependent Laser Filamentation Observables
- 1 December 2026
- Published byĀ Institute of Electrical and Electronics Engineers (IEEE)Ā inĀ IEEE Photonics Technology Letters
- Vol.Ā 38 Ā (23) , 1929-1932
Single-Image Inference of Power-Dependent Laser Filamentation Observables
- 9 November 2026
- Published byĀ SPIE-Intl Soc Optical Eng
- p. 14
Programmable wavefront control in coherent beam combination via deep learning
Coherent beam combination (CBC) is a promising technique to scale optical power far beyond the limits of a single fiber laser. In CBC, independent control over each channel allows the far-field intensity distribution to be reshaped, within the physical constraints of the system. Applications such as beam shaping, steering, and turbulence correction naturally benefit from this capability, as they require precisely the kind of wavefront adjustments that phase-only control provides. In practice, however, CBC systems are susceptible to phase noises, making real-time feedback control an essential requirement. Deep learning presents a promising approach, enabling single-step phase retrieval without iterative algorithms. In this work, we demonstrate a deep learning framework using a differential training strategy, which enables not only phase locking to maximize power-in-the-bucket, but also, flexible, programmable, and stable beam shaping and steering, including instantaneous intensity redistribution across target patterns. Additionally, we demonstrate its performance in the presence of turbulence. Altogether, this opens new possibilities for high-power laser systems requiring agile, real-time beam management.
- 9 November 2026
- Published byĀ SPIE-Intl Soc Optical Eng
- p. 14
Programmable wavefront control in coherent beam combination via deep learning
Coherent beam combination (CBC) is a promising technique to scale optical power far beyond the limits of a single fiber laser. In CBC, independent control over each channel allows the far-field intensity distribution to be reshaped, within the physical constraints of the system. Applications such as beam shaping, steering, and turbulence correction naturally benefit from this capability, as they require precisely the kind of wavefront adjustments that phase-only control provides. In practice, however, CBC systems are susceptible to phase noises, making real-time feedback control an essential requirement. Deep learning presents a promising approach, enabling single-step phase retrieval without iterative algorithms. In this work, we demonstrate a deep learning framework using a differential training strategy, which enables not only phase locking to maximize power-in-the-bucket, but also, flexible, programmable, and stable beam shaping and steering, including instantaneous intensity redistribution across target patterns. Additionally, we demonstrate its performance in the presence of turbulence. Altogether, this opens new possibilities for high-power laser systems requiring agile, real-time beam management.
- Open Access
- 5 November 2026
- Published byĀ Optica Publishing GroupĀ inĀ Optics Express
- Vol.Ā 34 Ā (23) , 43816
Physics-guided source selection for few-shot transfer learning to characterise bismuth-doped fiber amplifier
Bismuth-doped fiber amplifiers (BDFAs) can extend optical amplification beyond the C + L bands into the low-loss window of silica fibers, but each additional fiber normally requires extensive gain and noise-figure characterisation. Here, we demonstrate physics-guided source selection for few-shot transfer learning in Bi/Ge co-doped BDFAs. Four fibers spanning 5.4ā22.9 mol% GeO 2 were characterised over 480 operating points each. Across 12 directed transfer-learning pairs, a two-descriptor feature distance based on GeCl 4 flow during fiber preform fabrication and 1385-nm absorption predicted transfer benefit. Selecting the physically matched source reduced the target-data requirement by up to 7x, and prospective validation on an unseen fifth fiber correctly identified the optimal transfer-learning source. More broadly, these results show that source quality, as captured by simple physical descriptors, can be more important than source-data volume in photonic transfer learning.
- 5 November 2026
- Published byĀ Optica Publishing GroupĀ inĀ Optics Express
- Vol.Ā 34 Ā (23) , 43816
Open AccessPhysics-guided source selection for few-shot transfer learning to characterise bismuth-doped fiber amplifier
Bismuth-doped fiber amplifiers (BDFAs) can extend optical amplification beyond the C + L bands into the low-loss window of silica fibers, but each additional fiber normally requires extensive gain and noise-figure characterisation. Here, we demonstrate physics-guided source selection for few-shot transfer learning in Bi/Ge co-doped BDFAs. Four fibers spanning 5.4ā22.9 mol% GeO 2 were characterised over 480 operating points each. Across 12 directed transfer-learning pairs, a two-descriptor feature distance based on GeCl 4 flow during fiber preform fabrication and 1385-nm absorption predicted transfer benefit. Selecting the physically matched source reduced the target-data requirement by up to 7x, and prospective validation on an unseen fifth fiber correctly identified the optimal transfer-learning source. More broadly, these results show that source quality, as captured by simple physical descriptors, can be more important than source-data volume in photonic transfer learning.
- Open Access
- 12 September 2026
- Published byĀ American Chemical Society (ACS)Ā inĀ ACS Omega
- Vol.Ā 11 Ā (38) , 58085-58090
Deep-Learning-Assisted Single-Shot Plasma Emission Imaging for Pollen Morphology Reconstruction and Apparent Hydration-State Classification
Monitoring pollen is important for crop productivity, plant breeding, and environmental forecasting, yet current approaches remain limited in throughput and physiological insight. Here, we combine single-shot plasma emission imaging with artificial intelligence for pollen morphology reconstruction and apparent hydration state classification. Plasma plumes generated during laserāpollen interaction provide indirect measurements of pollen morphology, extending plasma-based analysis beyond conventional spectroscopic readouts. A conditional generative adversarial neural network reconstructed morphology from such plasma images, while chromaticity analysis revealed systematic color shifts between hydrated and dehydrated grains. A support vector machine using chromaticity histogram and scalar color features distinguished these groups with ā¼76% leave-one-out accuracy. The results suggest that plasma emission contains structural and hydration-associated information, supporting rapid pollen morphology inference and physiological phenotyping for agricultural and environmental monitoring.
- 12 September 2026
- Published byĀ American Chemical Society (ACS)Ā inĀ ACS Omega
- Vol.Ā 11 Ā (38) , 58085-58090
Open AccessDeep-Learning-Assisted Single-Shot Plasma Emission Imaging for Pollen Morphology Reconstruction and Apparent Hydration-State Classification
Monitoring pollen is important for crop productivity, plant breeding, and environmental forecasting, yet current approaches remain limited in throughput and physiological insight. Here, we combine single-shot plasma emission imaging with artificial intelligence for pollen morphology reconstruction and apparent hydration state classification. Plasma plumes generated during laserāpollen interaction provide indirect measurements of pollen morphology, extending plasma-based analysis beyond conventional spectroscopic readouts. A conditional generative adversarial neural network reconstructed morphology from such plasma images, while chromaticity analysis revealed systematic color shifts between hydrated and dehydrated grains. A support vector machine using chromaticity histogram and scalar color features distinguished these groups with ā¼76% leave-one-out accuracy. The results suggest that plasma emission contains structural and hydration-associated information, supporting rapid pollen morphology inference and physiological phenotyping for agricultural and environmental monitoring.
- Open Access
- 6 August 2026
- Published byĀ Optica Publishing GroupĀ inĀ Optics Continuum
- Vol.Ā 5 Ā (8) , 2686
Deep learning of beam-to-filament mapping and orientation in ultrafast laser filamentation
Ultrafast laser filamentation exhibits complex spatial and angular behaviour that is difficult to characterize in real time due to nonlinear propagation. Here, we investigate whether structured input beam geometry is reflected in the resulting filament plasma emission and can be analyzed using data-driven deep learning methods. Using femtosecond laser filaments generated from spatially structured beam masks over a range of rotations, we demonstrate two complementary mappings: (i) inference of beam rotation angle from side-view plasma emission images, and (ii) prediction of plasma emission patterns from structured beam profiles. On unseen beam masks, the image-to-image model achieves a mean SSIM of 0.8910, RMSE of 0.034, and PSNR of 30.1 dB, capturing the overall spatial structure and orientation of the plasma emission. While local discrepancies in fine features remain, the results show that structured variations in the input beam produce reproducible signatures in the observed plasma emission. These findings demonstrate that filament plasma emission retains measurable and learnable structure related to the input beam, enabling data-driven analysis of beamāfilament relationships.
- 6 August 2026
- Published byĀ Optica Publishing GroupĀ inĀ Optics Continuum
- Vol.Ā 5 Ā (8) , 2686
Open AccessDeep learning of beam-to-filament mapping and orientation in ultrafast laser filamentation
Ultrafast laser filamentation exhibits complex spatial and angular behaviour that is difficult to characterize in real time due to nonlinear propagation. Here, we investigate whether structured input beam geometry is reflected in the resulting filament plasma emission and can be analyzed using data-driven deep learning methods. Using femtosecond laser filaments generated from spatially structured beam masks over a range of rotations, we demonstrate two complementary mappings: (i) inference of beam rotation angle from side-view plasma emission images, and (ii) prediction of plasma emission patterns from structured beam profiles. On unseen beam masks, the image-to-image model achieves a mean SSIM of 0.8910, RMSE of 0.034, and PSNR of 30.1 dB, capturing the overall spatial structure and orientation of the plasma emission. While local discrepancies in fine features remain, the results show that structured variations in the input beam produce reproducible signatures in the observed plasma emission. These findings demonstrate that filament plasma emission retains measurable and learnable structure related to the input beam, enabling data-driven analysis of beamāfilament relationships.
- Open Access
- 29 June 2026
- Published byĀ Optica Publishing GroupĀ inĀ Optics Continuum
- Vol.Ā 5 Ā (7) , 2249
Enhancing lensless imaging through Fisher-information-engineered structured scattering
Single-shot lensless imaging is fundamentally limited by the Fisher information carried to the detector. Here, we demonstrate a simple way to increase this information by enhancing the scattered field using a high-spatial-frequency amplitude mask placed after a laser-illuminated object. This added modulation produces greater structured scattering and increases the per-photon Fisher information of the measurement. Using fungal spores as a test case, the grid-enhanced configuration yields richer diffraction signatures and higher Fisher information compared to the unmodulated case, with most samples above unity. Under identical training conditions, a neural network trained on grid-modulated data produces reconstructions that more closely match ground-truth microscope images. Saliency analysis shows that Fisher information-enhanced scattering allows the model to focus on first-order diffraction-ring features, improving both interpretability and stability. These results demonstrate that Fisher information-engineered structured scattering, implemented using a low-cost static mask, provides a general and hardware-efficient route to more informative and interpretable lensless imaging for computational microscopy and compact sensing systems.
- 29 June 2026
- Published byĀ Optica Publishing GroupĀ inĀ Optics Continuum
- Vol.Ā 5 Ā (7) , 2249
Open AccessEnhancing lensless imaging through Fisher-information-engineered structured scattering
Single-shot lensless imaging is fundamentally limited by the Fisher information carried to the detector. Here, we demonstrate a simple way to increase this information by enhancing the scattered field using a high-spatial-frequency amplitude mask placed after a laser-illuminated object. This added modulation produces greater structured scattering and increases the per-photon Fisher information of the measurement. Using fungal spores as a test case, the grid-enhanced configuration yields richer diffraction signatures and higher Fisher information compared to the unmodulated case, with most samples above unity. Under identical training conditions, a neural network trained on grid-modulated data produces reconstructions that more closely match ground-truth microscope images. Saliency analysis shows that Fisher information-enhanced scattering allows the model to focus on first-order diffraction-ring features, improving both interpretability and stability. These results demonstrate that Fisher information-engineered structured scattering, implemented using a low-cost static mask, provides a general and hardware-efficient route to more informative and interpretable lensless imaging for computational microscopy and compact sensing systems.
- Open Access
- 14 May 2026
- Published byĀ IOP PublishingĀ inĀ Machine Learning: Engineering
- Vol.Ā 2 Ā (1) , 015010
Deep learning-based reconstruction of irregular laser ablation crater geometry from single plasma projection
During ultrafast laser ablation, a plasma plume forms as the laser interacts with the material, while the resulting crater geometry reflects how energy is deposited. We present a machine learning framework that reconstructs the 2D ablation pattern using only a single side-view image of the laser-induced plasma. A conditional generative adversarial network is trained to map the plasma projection directly to the corresponding crater morphology, incorporating an edge-aware loss function to improve the reconstruction of irregular contours. Despite relying on a single projection, the model successfully recovers the dominant spatial structure of the ablation pattern. Analysis of the learned representations indicates that spatial features within the plasma image contain information beyond the primary viewing axis, enabling the reconstruction of laterally irregular geometries. While fine surface textures remain challenging to resolve from a single view, the method accurately predicts the dominant crater spatial morphology. The approach complements spectral techniques by providing morphological information from indirect imaging, supports real-time in situ monitoring, and offers a practical route toward predictive diagnostics in laser-material processing. The framework could be extended to multi view fusion potentially when higher reconstruction fidelity is required.
- 14 May 2026
- Published byĀ IOP PublishingĀ inĀ Machine Learning: Engineering
- Vol.Ā 2 Ā (1) , 015010
Open AccessDeep learning-based reconstruction of irregular laser ablation crater geometry from single plasma projection
During ultrafast laser ablation, a plasma plume forms as the laser interacts with the material, while the resulting crater geometry reflects how energy is deposited. We present a machine learning framework that reconstructs the 2D ablation pattern using only a single side-view image of the laser-induced plasma. A conditional generative adversarial network is trained to map the plasma projection directly to the corresponding crater morphology, incorporating an edge-aware loss function to improve the reconstruction of irregular contours. Despite relying on a single projection, the model successfully recovers the dominant spatial structure of the ablation pattern. Analysis of the learned representations indicates that spatial features within the plasma image contain information beyond the primary viewing axis, enabling the reconstruction of laterally irregular geometries. While fine surface textures remain challenging to resolve from a single view, the method accurately predicts the dominant crater spatial morphology. The approach complements spectral techniques by providing morphological information from indirect imaging, supports real-time in situ monitoring, and offers a practical route toward predictive diagnostics in laser-material processing. The framework could be extended to multi view fusion potentially when higher reconstruction fidelity is required.
- Open Access
- 30 April 2026
- Published byĀ Optica Publishing GroupĀ inĀ Optics Express
- Vol.Ā 34 Ā (9) , 17148-17166
Efficient phase identification in coherent beam combination using interpretable deep learning
Coherent beam combination (CBC) offers significant power scaling beyond the capabilities of individual fiber lasers, but its effectiveness is heavily reliant on precise phase stabilization. Recent advancements in deep learning have shown potential for phase retrieval from interference intensity patterns in a single step. However, the interpretability of deep learning models and the optimal positioning of the imaging system remain unresolved challenges. In this study, we employ a spatial light modulator to emulate a CBC system and systematically investigate the phase prediction accuracy at various axial positions of the imaging system. We demonstrate that phase-retrieval efficiency can be substantially enhanced by identifying regions of the interference pattern with higher phase sensitivity. This approach enables a significant reduction in input size, allowing for the use of a lightweight fully connected neural network to achieve a phase prediction error of approximately Ī»/60, with a pure fully connected neural network inference rate of approximately 35 kHz for a 7-beamlet hexagonal close-packed array.
- 30 April 2026
- Published byĀ Optica Publishing GroupĀ inĀ Optics Express
- Vol.Ā 34 Ā (9) , 17148-17166
Open AccessEfficient phase identification in coherent beam combination using interpretable deep learning
Coherent beam combination (CBC) offers significant power scaling beyond the capabilities of individual fiber lasers, but its effectiveness is heavily reliant on precise phase stabilization. Recent advancements in deep learning have shown potential for phase retrieval from interference intensity patterns in a single step. However, the interpretability of deep learning models and the optimal positioning of the imaging system remain unresolved challenges. In this study, we employ a spatial light modulator to emulate a CBC system and systematically investigate the phase prediction accuracy at various axial positions of the imaging system. We demonstrate that phase-retrieval efficiency can be substantially enhanced by identifying regions of the interference pattern with higher phase sensitivity. This approach enables a significant reduction in input size, allowing for the use of a lightweight fully connected neural network to achieve a phase prediction error of approximately Ī»/60, with a pure fully connected neural network inference rate of approximately 35 kHz for a 7-beamlet hexagonal close-packed array.
- Open Access
- 16 February 2026
- Published byĀ IOP PublishingĀ inĀ Journal of Physics: Photonics
- Vol.Ā 8 Ā (1) , 015058
Ultrafast laser filamentation classification and analysis via neural networks
Laser filamentation is a nonlinear propagation regime in which high-power ultrashort pulses self-guide through air, generating plasma emission and broadband spectra. It underpins applications ranging from remote sensing to lightning control, yet detecting the precise onset of filamentation remains challenging. Here, we demonstrate a deep learning approach for identification and analysis of filament formation. A convolutional neural network (CNN) achieved a higher accuracy in predicting onset directly from plasma emission images than a linear regression model. A complementary nonānegative matrix factorisationāCNN (NMFāCNN) regression revealed that spatial emission structure encodes sufficient information to reconstruct broadband spectra with strong fidelity (median R 2 = 0.953), linking image features to underlying physical processes. This methodology establishes a route toward real-time detection and analysis of ultrafast nonlinear lightāmatter interactions, with implications for laser diagnostics, high-power beam control, and photonic sensing.
- 16 February 2026
- Published byĀ IOP PublishingĀ inĀ Journal of Physics: Photonics
- Vol.Ā 8 Ā (1) , 015058
Open AccessUltrafast laser filamentation classification and analysis via neural networks
Laser filamentation is a nonlinear propagation regime in which high-power ultrashort pulses self-guide through air, generating plasma emission and broadband spectra. It underpins applications ranging from remote sensing to lightning control, yet detecting the precise onset of filamentation remains challenging. Here, we demonstrate a deep learning approach for identification and analysis of filament formation. A convolutional neural network (CNN) achieved a higher accuracy in predicting onset directly from plasma emission images than a linear regression model. A complementary nonānegative matrix factorisationāCNN (NMFāCNN) regression revealed that spatial emission structure encodes sufficient information to reconstruct broadband spectra with strong fidelity (median R 2 = 0.953), linking image features to underlying physical processes. This methodology establishes a route toward real-time detection and analysis of ultrafast nonlinear lightāmatter interactions, with implications for laser diagnostics, high-power beam control, and photonic sensing.
- Open Access
- 29 January 2026
- Published byĀ American Chemical Society (ACS)Ā inĀ ACS ES&T Air
- Vol.Ā 3 Ā (5) , 1243-1251
Airborne Particulate Matter Sensing via Laser FilamentāInteraction and Deep Learning
Airborne particulate matter, including mineral dust and biological aerosols, such as pollen, presents growing challenges to public health, agriculture, and environmental monitoring. Current technologies lack the capability to identify single particles remotely in real-time. Here, we demonstrate the feasibility of femtosecond laser filamentation combined with optical imaging and lightweight deep learning for the remote identification of airborne particulates. Chalk dust, pollen grains, and salt crystals were delivered into a laser-generated filament, where their optical emission events were captured and analyzed. A convolutional neural network trained on these events achieved high classification accuracy across all categories, with a mean accuracy of 87.5%. Grad-CAM visualizations confirmed that the network focused on discriminative spatial and spectral features for chalk, salt, and pollen. This study demonstrates the feasibility of remote, species-level airborne particulate detection and lays the foundation for intelligent, real-time atmospheric sensing platforms.
- 29 January 2026
- Published byĀ American Chemical Society (ACS)Ā inĀ ACS ES&T Air
- Vol.Ā 3 Ā (5) , 1243-1251
Open AccessAirborne Particulate Matter Sensing via Laser FilamentāInteraction and Deep Learning
Airborne particulate matter, including mineral dust and biological aerosols, such as pollen, presents growing challenges to public health, agriculture, and environmental monitoring. Current technologies lack the capability to identify single particles remotely in real-time. Here, we demonstrate the feasibility of femtosecond laser filamentation combined with optical imaging and lightweight deep learning for the remote identification of airborne particulates. Chalk dust, pollen grains, and salt crystals were delivered into a laser-generated filament, where their optical emission events were captured and analyzed. A convolutional neural network trained on these events achieved high classification accuracy across all categories, with a mean accuracy of 87.5%. Grad-CAM visualizations confirmed that the network focused on discriminative spatial and spectral features for chalk, salt, and pollen. This study demonstrates the feasibility of remote, species-level airborne particulate detection and lays the foundation for intelligent, real-time atmospheric sensing platforms.