Our research leverages deep learning to address complex inverse problems and pattern recognition tasks across interdisciplinary domains. Specifically, we employ convolutional neural network (CNN) architectures to enhance image processing capabilities and develop innovative solutions for challenges in medical diagnostics, engineering systems, and ultrafast science.
Linear-Chirp Instability Measurement of Ultrashort Laser Pulse Trains Using Convolutional Neural Networks
https://www.preprints.org/manuscript/202608.1129
We present a novel deep learning-based technique for measuring linear-chirp instability of ultrashort laser pulse trains directly from their Frequency-Resolved Optical Gating (FROG) traces. This approach requires neither an iterative retrieval algorithm nor pre- or post-processing steps, enabling real-time measurements. Although our proof-of-concept convolutional neural networks (CNNs) are moderately deep and trained on compact datasets, they demonstrate resilience to significant additive noise (up to 14%) and high robustness across a wide range of laser pulse parameter variations. When tested on unseen data, our classification network (ICN-1) achieved an accuracy of 80.0% (with an off-by-one accuracy of 100.0%), and our regression network (IRN-2) yielded a normalized root-mean-square error (NRMSE) of 10.2%.
Students in physics courses often encounter a persistent bottleneck that prevents them from making connections with the material and deeply understanding concepts: the overwhelmingly complex mathematical formalism. Students frequently become so bogged down in navigating sophisticated formulas and solving differential equations that their underlying physical intuition is lost. I integrate computational practices into my courses as a primary tool to address this issue, making physics more accessible — particularly for students with weaker mathematical backgrounds. Rather than treating computation as an extra assignment or a separate, disconnected coding lab, this approach positions computational tasks as a bridge that translates abstract mathematical pictures into a visible, interactive physical insights.
Fostering Conceptual Understanding Through Computational Practices in Introductory Modern Physics and Quantum Mechanics
https://planion.events/e/aapt/sm26/abstracts/52943/
Computational thinking has proven to be a powerful tool for enhancing students’ learning across various science disciplines. In this work, we discuss several examples of computational activities implemented in introductory modern physics and quantum mechanics courses. These activities are shown to improve students’ understanding of complex concepts and problems, such as wavepackets, Planck’s radiation law, Heisenberg's matrix mechanics, and finding wavefunctions for simple time-independent potentials.
Using MATLAB in Introductory Physics Courses: Enhanced Understanding of a Model and Addressing its Limitations
https://aapt.planion.com/Web.User/AbstractDet?ACCOUNT=AAPT&ABSID=20407&CONF=SM25&CKEY=
Projectile motion is one of the first topics discussed in introductory physics courses as an example of two-dimensional motion. Understanding how the separation of components works to provide a mathematical model for the projectile's trajectory based on the one-dimensional kinematic equations can be challenging for students. We present a concise MATLAB script that can be readily developed by students to visualize this and also examine the effect of the initial parameters. However, this simplified projectile motion model is often inadequate for real-world applications because it disregards factors such as air resistance and flow, curvature and rotation of the Earth, and variations of the gravitational field strength, among others. We show how MATLAB can empower students to explore and address these limitations through computational modeling, rather than relying solely on cumbersome analytical approaches.
Ultrashort laser pulses have been extensively employed to investigate transient phenomena occurring on the femtosecond and attosecond timescales. They have also been utilized in medicine and industry for surgical, imaging, and fabrication purposes, among many others. To understand and, more importantly, control the underlying ultrafast dynamics in various applications, knowledge of the exact electric field of the interacting pulse (i.e., its amplitude and phase) is often crucial. We develop novel computational techniques to fully characterize these laser pulses.
Reliable determination of pulse-shape instability in trains of ultrashort laser pulses using frequency-resolved optical gating
https://www.nature.com/articles/s41598-022-25193-3
We describe a reliable approach for determining the presence of pulse-shape instability in a train of ultrashort laser pulses. While frequency-resolved optical gating (FROG) has been shown to successfully perform this task by displaying a discrepancy between the measured and retrieved traces for unstable trains, it fails if its pulse-retrieval algorithm stagnates because algorithm stagnation and pulse-shape instability can be indistinguishable. So, a non-stagnating algorithm—even in the presence of instability—is required. The recently introduced Retrieved-Amplitude N-grid Algorithmic (RANA) approach has achieved extremely reliable (100%) pulse-retrieval in FROG for trains of stable pulse shapes, even in the presence of noise, and so is a promising candidate for an algorithm that can definitively distinguish stable and unstable pulse-shape trains. But it has not yet been considered for trains of pulses with pulse-shape instability. So, here, we investigate its performance for unstable trains of pulses with random pulse shapes. We consider trains of complex pulses measured by second-harmonic-generation FROG using the RANA approach and compare its performance to the well-known generalized-projections (GP) algorithm without the RANA enhancements. We show that the standard GP algorithm frequently fails to converge for such unstable pulse trains, yielding highly variable trace discrepancies. As a result, it is an unreliable indicator of instability. Using the RANA approach, on the other hand, we find zero stagnations, even for highly unstable pulse trains, and we conclude that FROG, coupled with the RANA approach, provides a highly reliable indicator of pulse-shape instability. It also provides a typical pulse length, spectral width, and time-bandwidth product, even in cases of instability.
Characterization of two-color ultrashort laser pulses using polarization-gating and transient-grating frequency-resolved optical gating
https://opg.optica.org/josab/abstract.cfm?uri=josab-39-3-683
Two-color ultrashort laser pulses have emerging applications in numerous areas of science and technology. In many cases, the slightest change in the combined electric field of a two-color pulse greatly affects its interaction mechanism with the system. Therefore, a precise characterization of the temporal/spectral profile of the combined electric field is of great importance. In this work, we demonstrate that a full characterization is possible using the well-known transient-grating (TG) or polarization-gating (PG) frequency-resolved optical gating (FROG) techniques, and by employing the recently developed Retrieved-Amplitude N-grid Algorithmic (RANA) approach for the retrieval process. We demonstrate the validity of using these techniques and this approach for multi-cycle and few-cycle pulses in the absence and presence of noise.