FIESTA: Multidisciplinary Applications in Research

• FIESTA is a polysemous technical label that designates diverse methods, systems, and software frameworks across fields like astrophysics, quantum field theory, robotics, and IoT.
• It enables advanced analyses ranging from Fourier-domain exoplanet spectroscopy that reduces noise to sector-decomposition algorithms for evaluating multiloop Feynman integrals.
• Its applications span incremental distance-field mapping in robotics and federated IoT platforms to immersive VR environments, highlighting both interdisciplinary impact and adaptability.

FIESTA is a polysemous technical label used across several research communities for unrelated methods, software systems, and platforms. In arXiv literature it denotes, among other things, FourIEr phase SpecTrum Analysis for exoplanet radial-velocity analysis (Zhao et al., 2022), the Feynman Integral Evaluation by a Sector decomposiTion Approach for multiloop quantum field theory (Tentyukov et al., 2010), Fast Incremental Euclidean DiSTAnce Fields for online motion planning (Han et al., 2019), and the Federated Interoperable Semantic IoT platform (Solmaz et al., 2019). The name also appears in immersive analytics, affective computing, Bayesian transient inference, tokamak equilibrium modeling, and medical imaging (Lee et al., 2020, Honarmand et al., 29 Mar 2025, Koehn et al., 18 Jul 2025, Pentland et al., 2024, Fadida et al., 2022).

1. Acronyms, capitalization, and semantic range

Across the cited works, the label appears in several orthographic forms and with distinct expansions. In exoplanet spectroscopy it is an improved FourIEr phase SpecTrum Analysis, also written as $\phi$ESTA (Zhao et al., 2022). In multiloop field theory it is the Feynman Integral Evaluation by a Sector decomposiTion Approach (Tentyukov et al., 2010). In aerial robotics it is Fast Incremental Euclidean DiSTAnce Fields (Han et al., 2019). In smart-city infrastructure it denotes the Federated Interoperable Semantic IoT platform (Solmaz et al., 2019). In immersive analytics it is the Free-roaming Immersive Environment to Support Team-based Analysis (Lee et al., 2020). In test-time adaptation for facial expression recognition it is Fisher Information-based Efficient Selective Test-time Adaptation (Honarmand et al., 29 Mar 2025).

The label also appears in mixed-case and lowercase variants. FiEstAS sampling is a Monte Carlo integration algorithm based on the Field Estimator for Arbitrary Spaces (0807.4479), while fiesta is a JAX-based Python package for Bayesian inference on kilonova and GRB afterglow lightcurves (Koehn et al., 18 Jul 2025). This breadth means that, in technical writing, “FIESTA” is not self-identifying and normally requires immediate domain qualification.

2. Exoplanet radial-velocity spectroscopy

In stellar radial-velocity work, FIESTA refers to an improved Fourier-domain method for separating true Doppler shifts from line-profile deformations in stellar spectra (Zhao et al., 2022). The method projects a stellar spectrum’s cross-correlation function onto a truncated set of Fourier basis functions and uses the amplitude and phase information from each Fourier mode to trace line variability at different cross-correlation-function width scales (Zhao et al., 2022). Its explicit aim is to disentangle apparent velocity shifts caused by stellar or instrumental line deformations from genuine Doppler motion.

The reported application is a three-year HARPS-N solar data set. In that study, $\phi$ESTA reveals the solar rotational effect, the long-term trend due to the solar magnetic cycle, instrumental instability, and apparent solar rotation-rate changes (Zhao et al., 2022). A multiple linear regression on $\phi$ESTA metrics reduces the weighted rms noise from $1.89\,\mathrm{m/s}$ to $0.98\,\mathrm{m/s}$, and the analysis also identifies an approximately three-day lag in the metrics, similar to earlier findings for BIS and FWHM (Zhao et al., 2022).

Within the exoplanet-detection context, FIESTA is therefore a Fourier-domain diagnostic and mitigation framework for activity- and instrument-induced contamination of Doppler measurements rather than a generic period-search or orbit-fitting method (Zhao et al., 2022).

3. Multiloop Feynman-integral evaluation

In perturbative quantum field theory, FIESTA is a long-running software family for the numerical evaluation of multiloop Feynman integrals via sector decomposition (Tentyukov et al., 2010). The core workflow is stable across releases: construct the Feynman-parameter representation, factorize UV and IR singularities by sector decomposition, perform the $\epsilon$-expansion, and numerically integrate the finite coefficients (Tentyukov et al., 2010). The Mathematica-facing interface is organized around calls such as SDEvaluate[UF[loop_momenta, propagators, subst], indices, order], with Cuba/VEGAS used in the early implementation examples (Tentyukov et al., 2010).

Successive releases broadened both functionality and scale. FIESTA 3 added cluster parallelization, new asymptotic expansion algorithms, and calculations in physical regions via contour deformation (Smirnov, 2013). FIESTA 4 targeted large-scale performance and introduced GPU support together with cluster-usage optimizations (Smirnov, 2015). FIESTA5 is explicitly performance-oriented, adds Quasi Monte Carlo and Tensor Train integrators, and rewrites most of the legacy code to the C++17 standard (Smirnov et al., 2021). These releases position FIESTA as one of the main public sector-decomposition engines for multiloop numerics.

The package also functions as a numerical backend in broader symbolic workflows. FeynCalc 10 and FeynHelpers expose FIESTA alongside FIRE, KIRA, pySecDec, and related tools, using Mathematica-side objects such as GLI and FCTopology and interface routines including FCLoopSwitchEtaSign, FSACreateMathematicaScripts, FSARunIntegration, and FSALoadNumericalResults (Shtabovenko, 2024, Shtabovenko, 22 Dec 2025). In that ecosystem, FIESTA occupies the post-IBP stage: master integrals are reduced symbolically and then evaluated numerically at chosen kinematic points.

4. Incremental distance fields and robotic planning

In robotics, FIESTA is a mapping system for building a global Euclidean Signed Distance Field incrementally for online motion planning of aerial robots (Han et al., 2019). Its design separates obstacle insertion and deletion into two independent updating queues, uses indexing data structures and doubly linked lists for map maintenance, and updates as few nodes as possible within a BFS framework (Han et al., 2019). The stated purpose is real-time ESDF maintenance with high computational performance and near-optimal results, and the method is integrated into a completed quadrotor system and validated in simulation and onboard experiments (Han et al., 2019).

A later whole-body planning study treats FIESTA as one of three state-of-the-art distance-field back ends, together with Voxblox and GPU-Voxels (Finean et al., 2021). In that comparison, FIESTA is described as a CPU-based framework that incrementally updates an occupancy-grid distance field and can be adapted to produce signed distance fields by calculating distance fields for both the occupancy map and the inverse occupancy map and subtracting them (Finean et al., 2021). The integrated hybrid system in that paper ultimately uses GPU-Voxels and GPMP2, but FIESTA serves as a benchmarked reference point for real-time reconstruction in dynamic environments (Finean et al., 2021).

This usage anchors FIESTA in the distance-field branch of motion planning, where the key commodity is not a map per se but a continually updated field that supports distance and gradient queries for trajectory optimization (Han et al., 2019, Finean et al., 2021).

5. Federated infrastructures and immersive collaborative systems

In smart-city and IoT research, FIESTA appears as FIESTA-IoT, the Federated Interoperable Semantic IoT platform (Solmaz et al., 2019). It is described as a federated interoperable semantic IoT platform and as a “system of systems,” intended to integrate heterogeneous testbeds and sensor infrastructures through a common semantic layer (Solmaz et al., 2019). The paper applies this framework to crowd mobility, combining a Crowd Mobility Analytics System and a Crowd Counting and Location System in pilots conducted in Gold Coast, Australia, and Santander, Spain (Solmaz et al., 2019). In this usage, FIESTA designates a semantic federation and interoperability substrate rather than an inference algorithm.

In immersive analytics, FIESTA denotes the Free-roaming Immersive Environment to Support Team-based Analysis, a prototype multi-user VR system for co-located collaborative data analysis (Lee et al., 2020). Users can freely position authoring interfaces and visualization artefacts anywhere in the virtual environment, either on virtual surfaces or suspended within the interaction space (Lee et al., 2020). A study with teams of three participants solving visual analytics tasks reports that wall surfaces were often used to organize 2D visualizations, while 3D visualizations were positioned in the surrounding space, and that outside tightly coupled collaboration participants generally did not manipulate visualizations that belonged to others (Lee et al., 2020).

These two systems are unrelated in purpose but share a structural emphasis on coordination across heterogeneous agents or viewpoints: one across distributed IoT infrastructures, the other across co-located analysts in a shared immersive workspace (Solmaz et al., 2019, Lee et al., 2020).

6. Additional domain-specific uses

Beyond the better-known meanings above, the term is attached to several other technically specific systems, packages, and sequences. The table summarizes representative uses documented in the cited records.

Usage	Field	Description
FIESTA	Medical image segmentation	Fourier-Based Semantic Augmentation with Uncertainty Guidance for Enhanced Domain Generalizability in Medical Image Segmentation; the abstract describes amplitude and phase manipulation in the frequency domain with epistemic-uncertainty guidance for single-source domain generalization (Oh et al., 2024)
FIESTA	Test-time adaptation	Fisher Information-based Efficient Selective Test-time Adaptation for video-based facial expression recognition; on AffWild2 it reports a $7.7\%$ improvement in F1 score over the base model while adapting only $22{,}000$ parameters (Honarmand et al., 29 Mar 2025)
fiesta	Time-domain astrophysics	A JAX-based Python package for training ML surrogates for GRB afterglow and kilonova models and for posterior sampling with flowMC; it supports GPU acceleration in surrogate training and sampling (Koehn et al., 18 Jul 2025)
Fiesta	Tokamak equilibrium modeling	A MATLAB-based free-boundary static equilibrium solver that can carry out forward and inverse equilibrium calculations and linearised dynamic modeling using the RZIp rigid plasma framework (Pentland et al., 2024)
FIESTA	Model selection	Fast IdEntification of State-of-The-Art models, a bandit-based model selection framework that adaptively allocates train-test splits and random seeds to candidate models (Moss et al., 2019)
FiEstAS sampling	Numerical integration	A Monte Carlo algorithm for multidimensional numerical integration based on the Field Estimator for Arbitrary Spaces, with emphasis on multimodal distributions and strong parameter degeneracies (0807.4479)
Fiesta	Network security	A web/HTTP exploitation kit discussed together with Mpack and ICEpack as an “enhanced IRC tool” in an encapsulated botnet-detection architecture (Osagie et al., 2019)
FIESTA	Fetal MRI	Fast Imaging Employing Steady-State Acquisition, one of the two fetal MRI sequences in a partial-annotation study; on out-of-distribution FIESTA data, Dice improves from $0.84$ to $0.9$ and ASSD decreases from $\phi$0 to $\phi$1 mm under the reported partial-annotation regime (Fadida et al., 2022)

This distribution suggests that FIESTA functions less as a stable field-specific term than as a reusable project name whose meaning is determined entirely by disciplinary context. In practice, precise interpretation depends on the accompanying expansion, capitalization, or software ecosystem.