LASED: Laser-Assisted Dynamics and Datasets

• LASED is a multifaceted term defining laser-assisted electron scattering, Auger decay, quantum dynamics simulations, and benchmark ML datasets with applications in physics and remote sensing.
• It details theoretical models including Volkov states and photon-order interference, emphasizing experimental observables like resonant peaks and angular distributions.
• LASED also offers simulation tools and curated datasets that drive advancements in sensor signal enhancement and visual place recognition, bridging physics and machine learning.

LASED encompasses a set of distinct entities in atomic physics, quantum simulations, sensor signal processing, and remote sensing. Across these domains, "LASED" denotes either methodologies (laser-assisted electron scattering, laser-enabled Auger decay), specialized simulators (quantum electrodynamics-based code), or large-scale datasets for machine learning and optimization (accelerometry, visual place recognition). The following organizes the principal variants, formal definitions, methodologies, datasets, and practical implications with reference to the primary literature.

1. Laser-Assisted Electron-Atom Scattering (LASED)

The term LASED is frequently used as shorthand for laser-assisted electron-atom scattering, most notably in the theoretical description of inelastic electron-hydrogen scattering in a monochromatic, linearly polarized laser field (Buica, 2023). The process $e^-(p_i) + H(2s) \rightarrow e^-(p_f) + H(n\ell)$ is analyzed under high-energy and moderate-intensity conditions.

Theoretical Formalism

• Volkov States: The projectile electron in the external field is modeled as a Gordon–Volkov state,

$$\psi_p(\mathbf{r}, t) = (2\pi)^{-3/2} \exp\left\{ i \left[ \mathbf{p}\cdot\mathbf{r} - E_p t/\hbar - \alpha_p \sin(\omega t) \right] \right\}$$

with $E_p = p^2/(2m)$ and $\alpha_p = (e \varepsilon_0 \cdot p) / (m\omega^2)$.

• Transition Amplitude: In the first Born approximation and to first order in atomic dressing,

$$M_{fi}^{(n)} = \int d^3r \int_{-\infty}^{+\infty} dt\, \psi^*_{p_f}(\mathbf{r}, t) V_{\text{scatt}}(\mathbf{r}) \psi_{p_i}(\mathbf{r}, t) e^{+i n \omega t}$$

Here $$V_{\text{scatt}}(\mathbf{r}) = -e^2/(4\pi\varepsilon_0)|\mathbf{r}|^{-1}$$.

• Laser-Assisted Differential Cross Section: For the $n$-photon channel,

$$\frac{d\sigma_n}{d\Omega} = (2\pi)^4 \frac{p_f}{p_i} |M_{fi}^{(n)}|^2 \delta(E_i - E_f - n\hbar\omega)$$

Energy conservation incorporates the net absorption/emission of $n$ photons.

• Photon-Order Structure: The interference of Volkov phases leads to a Bessel function distribution,

$$\exp[-i\alpha_i \sin(\omega t)] \exp[+i\alpha_f \sin(\omega t)] = \exp[-i(\alpha_i-\alpha_f)\sin(\omega t)] = \sum_n J_n(\beta) e^{-in\omega t}$$

with $$\psi_p(\mathbf{r}, t) = (2\pi)^{-3/2} \exp\left\{ i \left[ \mathbf{p}\cdot\mathbf{r} - E_p t/\hbar - \alpha_p \sin(\omega t) \right] \right\}$$0.

• Explicit Transition Matrix Element:

$$\psi_p(\mathbf{r}, t) = (2\pi)^{-3/2} \exp\left\{ i \left[ \mathbf{p}\cdot\mathbf{r} - E_p t/\hbar - \alpha_p \sin(\omega t) \right] \right\}$$1

• $$\psi_p(\mathbf{r}, t) = (2\pi)^{-3/2} \exp\left\{ i \left[ \mathbf{p}\cdot\mathbf{r} - E_p t/\hbar - \alpha_p \sin(\omega t) \right] \right\}$$2.

Observables and Predicted Phenomenology

• Small-Angle Dominance: $$\psi_p(\mathbf{r}, t) = (2\pi)^{-3/2} \exp\left\{ i \left[ \mathbf{p}\cdot\mathbf{r} - E_p t/\hbar - \alpha_p \sin(\omega t) \right] \right\}$$3 is forward-peaked due to the behavior of $$\psi_p(\mathbf{r}, t) = (2\pi)^{-3/2} \exp\left\{ i \left[ \mathbf{p}\cdot\mathbf{r} - E_p t/\hbar - \alpha_p \sin(\omega t) \right] \right\}$$4 at low $$\psi_p(\mathbf{r}, t) = (2\pi)^{-3/2} \exp\left\{ i \left[ \mathbf{p}\cdot\mathbf{r} - E_p t/\hbar - \alpha_p \sin(\omega t) \right] \right\}$$5.
• Resonances: For $$\psi_p(\mathbf{r}, t) = (2\pi)^{-3/2} \exp\left\{ i \left[ \mathbf{p}\cdot\mathbf{r} - E_p t/\hbar - \alpha_p \sin(\omega t) \right] \right\}$$6, the matrix element diverges, producing sharp peaks.
• Photon Sidebands: Channels $$\psi_p(\mathbf{r}, t) = (2\pi)^{-3/2} \exp\left\{ i \left[ \mathbf{p}\cdot\mathbf{r} - E_p t/\hbar - \alpha_p \sin(\omega t) \right] \right\}$$7 dominate when $$\psi_p(\mathbf{r}, t) = (2\pi)^{-3/2} \exp\left\{ i \left[ \mathbf{p}\cdot\mathbf{r} - E_p t/\hbar - \alpha_p \sin(\omega t) \right] \right\}$$8.
• Nodal Dips: At $$\psi_p(\mathbf{r}, t) = (2\pi)^{-3/2} \exp\left\{ i \left[ \mathbf{p}\cdot\mathbf{r} - E_p t/\hbar - \alpha_p \sin(\omega t) \right] \right\}$$9, the cross section exhibits angular minima.
• Experimental Regime: $E_p = p^2/(2m)$0 W/cm$E_p = p^2/(2m)$1, $E_p = p^2/(2m)$2–5 eV, $E_p = p^2/(2m)$3–1000 eV.

2. Laser-Enabled Auger Decay (LEAD and spLEAD)

LASED is also cited in the context of Laser-Enabled Auger Decay (LEAD), notably single-photon variants (spLEAD) in atomic ions (Iablonskyi et al., 2017). This mechanism enables otherwise forbidden Auger decay channels by the absorption of photons.

Process and Amplitude

• Mechanism: An inner-valence hole in an ion absorbs a photon ($E_p = p^2/(2m)$4) and, via electronic correlation, decays to a dication plus a free electron—even when field-free decay is energetically forbidden.
• Quantum Amplitude:

$E_p = p^2/(2m)$5

• Decay Rate: By Fermi’s golden rule,

$E_p = p^2/(2m)$6

• Interference: In a bichromatic field, spLEAD ($E_p = p^2/(2m)$7-wave) and direct $E_p = p^2/(2m)$8 photoionization ($E_p = p^2/(2m)$9-wave) coherently interfere, yielding modulated photoelectron angular distributions.

Experimental and Conceptual Significance

• Correlation Sensitivity: spLEAD amplitude vanishes in the independent-particle model; strong dependence on electron correlation.
• Detection: Quantum-path interference is exploited to detect and control otherwise subdominant decay channels.
• Applications: Potential for probing ultrafast charge migration in molecules; cross-sections for molecular systems (e.g., glycine) predicted much higher than neon.

3. LASED: Laser-Atom Interaction Simulator

LASED also denotes a Python-based numerical toolbox for simulating laser-atom interaction dynamics from first QED principles (Patel et al., 2022). This implementation systematically models arbitrary laser polarization, Gaussian spatial profiles, Doppler broadening, and atomic substructure.

Hamiltonian and Quantum Dynamics

• Full Hamiltonian:

$\alpha_p = (e \varepsilon_0 \cdot p) / (m\omega^2)$0

with $\alpha_p = (e \varepsilon_0 \cdot p) / (m\omega^2)$1 (atomic eigenstates), $\alpha_p = (e \varepsilon_0 \cdot p) / (m\omega^2)$2 (quantized modes), and $\alpha_p = (e \varepsilon_0 \cdot p) / (m\omega^2)$3 (dipolar coupling).

• Density Matrix Formalism: Evolution governed by coupled equations for populations and coherences, including detuning, Rabi frequencies, and spontaneous decay.
• Laser Parameters: Arbitrary polarization in the $\alpha_p = (e \varepsilon_0 \cdot p) / (m\omega^2)$4 basis, intensity profiles (Gaussian TEM$\alpha_p = (e \varepsilon_0 \cdot p) / (m\omega^2)$5), spatial averaging over radial zones, and reference-frame rotation via Wigner D-matrices.

Numerical Workflow

• State Definition: User designates ground and excited state manifolds with full quantum numbers.
• Beam Profile: Gaussian intensity modeled by discrete rings; coherent dynamics tracked per ring and summed.
• Doppler Averaging: Maxwell-Boltzmann velocity distribution sampled, density matrices weighted and averaged.
• Time Evolution: Linear system $\alpha_p = (e \varepsilon_0 \cdot p) / (m\omega^2)$6 solved via diagonalization ($\alpha_p = (e \varepsilon_0 \cdot p) / (m\omega^2)$7).

Example Applications

• Atomic Systems: Simulations for Ca $\alpha_p = (e \varepsilon_0 \cdot p) / (m\omega^2)$8, He $\alpha_p = (e \varepsilon_0 \cdot p) / (m\omega^2)$9 post electron-impact excitation, Cs $$M_{fi}^{(n)} = \int d^3r \int_{-\infty}^{+\infty} dt\, \psi^*_{p_f}(\mathbf{r}, t) V_{\text{scatt}}(\mathbf{r}) \psi_{p_i}(\mathbf{r}, t) e^{+i n \omega t}$$0 hyperfine structure.
• Physical Insights: Predicts optical pumping, Rabi oscillations, substate population dynamics, and realistic Doppler/beam profile effects.

4. LASED Datasets in Machine Learning

Separate from atomic and quantum physics, LASED identifies two major datasets in machine learning and optimization: one for sensor enhancement (accelerometry), the other for visual place recognition from aerial imagery.

(A) Low-cost Accelerometer Signal Enhancement Dataset

• Purpose: First public dataset to support signal enhancement in low-cost accelerometry (range extension, noise reduction) (Wang et al., 25 Feb 2025).
• Composition:
  • $$M_{fi}^{(n)} = \int d^3r \int_{-\infty}^{+\infty} dt\, \psi^*_{p_f}(\mathbf{r}, t) V_{\text{scatt}}(\mathbf{r}) \psi_{p_i}(\mathbf{r}, t) e^{+i n \omega t}$$1Tens of thousands of tri-axial accelerometer records from 10 MEMS-equipped smartphones.
  • Training set: HONOR Magic 4; test: 9 additional models (including iPhone and major Android lines).
  • Motion scenarios: vigorous multiaxial hand-shaking to induce sensor saturation.
  • Reference: Trajectories captured by external 8-camera optical motion-capture (not co-mounted precision IMUs).
• Organization: Unpaired structure; training and test segments strictly partitioned by device.
• **File and metadata conventions, sample rates, and access details are not specified in the source paper.

(B) Large-Scale Estonia vPR Dataset

• Purpose: To address gaps in UAV visual place recognition, LASED (LArge-Scale Estonia vPR Dataset) presents nearly 1 million circular aerial images across $$M_{fi}^{(n)} = \int d^3r \int_{-\infty}^{+\infty} dt\, \psi^*_{p_f}(\mathbf{r}, t) V_{\text{scatt}}(\mathbf{r}) \psi_{p_i}(\mathbf{r}, t) e^{+i n \omega t}$$2170,000 non-overlapping locations in Estonia over 10 years (Papapetros et al., 20 Jul 2025).
• Sampling Regimen:
  • Circular (diameter 400 m / 500 px) orthoimages sampled yearly (2011–2021) with explicit place separation—geodesic distance between any two centers $$M_{fi}^{(n)} = \int d^3r \int_{-\infty}^{+\infty} dt\, \psi^*_{p_f}(\mathbf{r}, t) V_{\text{scatt}}(\mathbf{r}) \psi_{p_i}(\mathbf{r}, t) e^{+i n \omega t}$$3 400 m.
  • Each location paired across years via pixel-to-pixel correspondence.

Dataset	Domain	Scale / Size	Main Application
LASED (Physics)	Atoms/QED	-	Laser-atom dynamics simulation
LASED (ML/Acc.)	Sensor ML	Tens of thousands segs	Signal enhancement GANs
LASED (vPR)	Remote sens	$$M_{fi}^{(n)} = \int d^3r \int_{-\infty}^{+\infty} dt\, \psi^*_{p_f}(\mathbf{r}, t) V_{\text{scatt}}(\mathbf{r}) \psi_{p_i}(\mathbf{r}, t) e^{+i n \omega t}$$41M images	Place recognition (UAV)

5. Evaluation, Metrics, and Benchmarks

LASED in Physics

• Laser-assisted cross sections: Theoretical predictions for $$M_{fi}^{(n)} = \int d^3r \int_{-\infty}^{+\infty} dt\, \psi^*_{p_f}(\mathbf{r}, t) V_{\text{scatt}}(\mathbf{r}) \psi_{p_i}(\mathbf{r}, t) e^{+i n \omega t}$$5 have been calculated with explicit Bessel-function photon sideband weights. Resonant structure and small-angle peaks are direct consequences of atomic and laser parameters (Buica, 2023).

LASED in Machine Learning

• Accelerometer Signal Enhancement:
  • GAN-based approaches leveraging LASED with optimal transport supervision (OTS) and modulated Laplace energy (MLE) regularization can double dynamic range and reduce noise by two orders of magnitude compared to prior SOTA methods.
  • OTS defines the cost as $$M_{fi}^{(n)} = \int d^3r \int_{-\infty}^{+\infty} dt\, \psi^*_{p_f}(\mathbf{r}, t) V_{\text{scatt}}(\mathbf{r}) \psi_{p_i}(\mathbf{r}, t) e^{+i n \omega t}$$6 with optimal matching via transport plan $$M_{fi}^{(n)} = \int d^3r \int_{-\infty}^{+\infty} dt\, \psi^*_{p_f}(\mathbf{r}, t) V_{\text{scatt}}(\mathbf{r}) \psi_{p_i}(\mathbf{r}, t) e^{+i n \omega t}$$7.
  • Benchmarks outside enhancement results are not reported; no baseline Allan-variance, nor raw distribution figures are provided.
• Visual Place Recognition:
  • Models trained on LASED achieve average Recall@1 of $$M_{fi}^{(n)} = \int d^3r \int_{-\infty}^{+\infty} dt\, \psi^*_{p_f}(\mathbf{r}, t) V_{\text{scatt}}(\mathbf{r}) \psi_{p_i}(\mathbf{r}, t) e^{+i n \omega t}$$845% (ResNet50+CosPlace) versus 17–28% for smaller sets, with steerable CNNs (sResNet50c8) reaching 64.5% Recall@1.
  • Cross-dataset generalization is demonstrated, and models show enhanced robustness to both geographic and rotation-induced visual variation.

6. Practical Implications and Use Cases

• Laser-assisted scattering experiments: Observable, tunable angular and photon sideband structures in electron-atom collision cross sections are predicted as laser intensity and photon energy are varied (Buica, 2023).
• spLEAD detection: Pathway interference in bichromatic fields enables phase-resolved control and detection of weakly allowed Auger decay, opening new probes for electron correlation and attosecond hole dynamics (Iablonskyi et al., 2017).
• LASED Python simulation: Used for precise simulation of atomic state population evolution under realistic, user-specified laser fields and atomic structure, incorporating Gaussian, polarization, and temperature effects (Patel et al., 2022).
• Accelerometer benchmarking and ML: LASED (signal dataset) sets a new benchmark for sensor enhancement enabling GAN-based architectures to reliably "upgrade" cheap MEMS IMU readings, facilitating cost-effective large-scale deployment in wearables and robotics (Wang et al., 25 Feb 2025).
• Global-scale aerial vPR: LASED (vPR dataset) is foundational for robust UAV geolocalization in GNSS-denied environments, especially when using steerable CNNs to natively handle image rotation and maximize long-term retrieval accuracy (Papapetros et al., 20 Jul 2025).

7. Limitations and Unspecified Elements

• Several LASED variants (especially datasets) lack complete disclosure of sample rates, file formats, metadata conventions, calibration procedures, and access instructions in their primary sources (Wang et al., 25 Feb 2025, Papapetros et al., 20 Jul 2025).
• For physics-based LASED, predictions pertain primarily to hydrogen-like or simple atomic systems in high-field or low-density regimes, and generalization to complex atoms or molecules is implied but not detailed.
• All machine-learning datasets are single-country (Estonia) or smartphone-based, which may limit global or hardware-agnostic generalization.

---

For each domain, LASED thus designates foundational theory, computational tools, and datasets that set benchmarks or enable new experimental/algorithmic paradigms in their respective fields (Buica, 2023, Iablonskyi et al., 2017, Patel et al., 2022, Wang et al., 25 Feb 2025, Papapetros et al., 20 Jul 2025).