Vlad Gaciu

Vlad Gaciu

Auto-generated profile, based on publicly available information

Vlad Gaciu

‘Horia Hulubei’ National Institute for Physics and Nuclear Engineering

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 April 2024 in AIP Advances
Verified
Journal Article
Open access
PDF file
VGV. Gaciu
V. Gaciu
IDIoan Dancus
Ioan Dancus
BDB. Diaconescu
B. Diaconescu
DGD. G. Ghiţă
D. G. Ghiţă
ESEmil Slusanschi
Emil Slusanschi
CTCătălin Mihai Ticos
Cătălin Mihai Ticos
The high power laser system at Extreme Light Infrastructure—Nuclear Physics has demonstrated 10 PW power shot capability. It can also deliver beams with powers of 1 PW and 100 TW in several different experimental areas that carry out dedicated sets of experiments. An array of diagnostics is deployed to characterize the laser beam spatial profiles and to monitor their evolution during the amplification stages. Some of the essential near-field and far-field profiles acquired with CCD cameras are monitored constantly on a large screen television for visual observation and for decision making concerning the control and tuning of the laser beams. Here, we present results on the beam profile classification obtained from datasets with over 14 600 near-field and far-field images acquired during two days of laser operation at 1 PW and 100 TW. We utilize supervised and unsupervised machine learning models based on trained neural networks and an autoencoder. These results constitute an early demonstration of machine learning being used as a tool in the laser system data classification.

1 June 2026 in Machine Learning: Science and Technology
Verified
Journal Article
Open access
PDF file
VGVlad Gaciu
Vlad Gaciu
IDIoan Dancus
Ioan Dancus
MCMihai Caragea
Mihai Caragea
AGAndrei Grădinariu
Andrei Grădinariu
Alice Diana DumitruAlice Diana Dumitru
Alice Diana Dumitru
CTCătălin Mihai Ticos
Cătălin Mihai Ticos
We present a hybrid self-supervised model that integrates a convolutional autoencoder with a Guassian mixture model (GMM) to classify and reconstruct real petawatt laser beam profiles. The model distinguishes between near-field, far-field, and multi-spot beam profiles, demonstrating that self-supervised representations can extract physically meaningful features from unlabeled experimental laser data sets. We achieve 89.4 % classification accuracy on our baseline model, improving the previously unsupervised baseline of 85 %, under comparable experimental conditions. The integration of a probabilistic layer GMM enables more precise clustering of laser beam profiles, achieving 98.9% classification accuracy. Designed for an inference speed up to 1 Hz for live diagnostics on the ELI-NP high power laser system, our architecture enables real-time state detection. Importantly, we show that high classification performance can emerge even when pixel-level reconstructions are imperfect, indicating that the learned latent representations prioritize physically meaningful structure over image fidelity.