arXiv subjects

Jose L. Gómez

Publications and source records attributed to Jose L. Gómez.

11 recordsLinked to original sources

arXiv2026

FedS2R: One-Shot Federated Domain Generalization for Synthetic-to-Real Semantic Segmentation in Autonomous Driving

Federated domain generalization has shown promising progress in image classification by enabling collaborative training across multiple clients without sharing raw data. However, its potential in the semantic segmentation of autonomous driving remains underexplored. In this paper, we propose FedS2R, the first one-shot federated domain generalization framework for synthetic-to-real semantic segmentation in autonomous driving. FedS2R comprises two components: an inconsistency-driven data augmentation strategy that generates images for unstable classes, and a multi-client knowledge distillation scheme with feature fusion that distills a global model from multiple client models. Experiments on five real-world datasets, Cityscapes, BDD100K, Mapillary, IDD, and ACDC, show that the global model significantly outperforms individual client models and is only 2 mIoU points behind the model trained with simultaneous access to all client data. These results demonstrate the effectiveness of FedS2R in synthetic-to-real semantic segmentation for autonomous driving under federated learning

cs.CV↗

arXiv2025

All for One, and One for All: UrbanSyn Dataset, the third Musketeer of Synthetic Driving Scenes

We introduce UrbanSyn, a photorealistic dataset acquired through semi-procedurally generated synthetic urban driving scenarios. Developed using high-quality geometry and materials, UrbanSyn provides pixel-level ground truth, including depth, semantic segmentation, and instance segmentation with object bounding boxes and occlusion degree. It complements GTAV and Synscapes datasets to form what we coin as the 'Three Musketeers'. We demonstrate the value of the Three Musketeers in unsupervised domain adaptation for image semantic segmentation. Results on real-world datasets, Cityscapes, Mapillary Vistas, and BDD100K, establish new benchmarks, largely attributed to UrbanSyn. We make UrbanSyn openly and freely accessible (www.urbansyn.org).

cs.CV↗

arXiv2024

Divide, Ensemble and Conquer: The Last Mile on Unsupervised Domain Adaptation for Semantic Segmentation

The last mile of unsupervised domain adaptation (UDA) for semantic segmentation is the challenge of solving the syn-to-real domain gap. Recent UDA methods have progressed significantly, yet they often rely on strategies customized for synthetic single-source datasets (e.g., GTA5), which limits their generalisation to multi-source datasets. Conversely, synthetic multi-source datasets hold promise for advancing the last mile of UDA but remain underutilized in current research. Thus, we propose DEC, a flexible UDA framework for multi-source datasets. Following a divide-and-conquer strategy, DEC simplifies the task by categorizing semantic classes, training models for each category, and fusing their outputs by an ensemble model trained exclusively on synthetic datasets to obtain the final segmentation mask. DEC can integrate with existing UDA methods, achieving state-of-the-art performance on Cityscapes, BDD100K, and Mapillary Vistas, significantly narrowing the syn-to-real domain gap.

cs.CV↗

arXiv2023

Co-Training for Unsupervised Domain Adaptation of Semantic Segmentation Models

Semantic image segmentation is a central and challenging task in autonomous driving, addressed by training deep models. Since this training draws to a curse of human-based image labeling, using synthetic images with automatically generated labels together with unlabeled real-world images is a promising alternative. This implies to address an unsupervised domain adaptation (UDA) problem. In this paper, we propose a new co-training procedure for synth-to-real UDA of semantic segmentation models. It consists of a self-training stage, which provides two domain-adapted models, and a model collaboration loop for the mutual improvement of these two models. These models are then used to provide the final semantic segmentation labels (pseudo-labels) for the real-world images. The overall procedure treats the deep models as black boxes and drives their collaboration at the level of pseudo-labeled target images, i.e., neither modifying loss functions is required, nor explicit feature alignment. We test our proposal on standard synthetic and real-world datasets for on-board semantic segmentation. Our procedure shows improvements ranging from ~13 to ~26 mIoU points over baselines, so establishing new state-of-the-art results.

cs.CV↗

arXiv2021

Probing the innermost regions of AGN jets and their magnetic fields with RadioAstron. V. Space and ground millimeter-VLBI imaging of OJ 287

We present the first polarimetric space VLBI observations of OJ 287, observed with RadioAstron at 22 GHz during a perigee session on 2014 April 4 and five near-in-time snapshots, together with contemporaneous ground VLBI observations at 15, 43, and 86 GHz. Ground-space fringes were obtained up to a projected baseline of 3.9 Earth diameters during the perigee session, and at a record 15.1 Earth diameters during the snapshot sessions, allowing us to image the innermost jet at an angular resolution of $\sim50μ$as, the highest ever achieved at 22 GHz for OJ 287. Comparison with ground-based VLBI observations reveals a progressive jet bending with increasing angular resolution that agrees with predictions from a supermassive binary black hole model, although other models cannot be ruled out. Spectral analyses suggest that the VLBI core is dominated by the internal energy of the emitting particles during the onset of a multi-wavelength flare, while the parsec-scale jet is consistent with being in equipartition between the particles and magnetic field. Estimated minimum brightness temperatures from the visibility amplitudes show a continued rising trend with projected baseline length up to $10^{13}$ K, reconciled with the inverse Compton limit through Doppler boosting for a jet closely oriented to the line of sight. The observed electric vector position angle suggests that the innermost jet has a predominantly toroidal magnetic field, which together with marginal evidence of a gradient in rotation measure across the jet width indicate that the VLBI core is threaded by a helical magnetic field, in agreement with jet formation models.

astro-ph.HE↗

arXiv2021

The Variability of the Black-Hole Image in M87 at the Dynamical Time Scale

The black-hole images obtained with the Event Horizon Telescope (EHT) are expected to be variable at the dynamical timescale near their horizons. For the black hole at the center of the M87 galaxy, this timescale (5-61 days) is comparable to the 6-day extent of the 2017 EHT observations. Closure phases along baseline triangles are robust interferometric observables that are sensitive to the expected structural changes of the images but are free of station-based atmospheric and instrumental errors. We explored the day-to-day variability in closure phase measurements on all six linearly independent non-trivial baseline triangles that can be formed from the 2017 observations. We showed that three triangles exhibit very low day-to-day variability, with a dispersion of $\sim3-5^\circ$. The only triangles that exhibit substantially higher variability ($\sim90-180^\circ$) are the ones with baselines that cross visibility amplitude minima on the $u-v$ plane, as expected from theoretical modeling. We used two sets of General Relativistic magnetohydrodynamic simulations to explore the dependence of the predicted variability on various black-hole and accretion-flow parameters. We found that changing the magnetic field configuration, electron temperature model, or black-hole spin has a marginal effect on the model consistency with the observed level of variability. On the other hand, the most discriminating image characteristic of models is the fractional width of the bright ring of emission. Models that best reproduce the observed small level of variability are characterized by thin ring-like images with structures dominated by gravitational lensing effects and thus least affected by turbulence in the accreting plasmas.

astro-ph.HE↗

arXiv2021

Co-training for Deep Object Detection: Comparing Single-modal and Multi-modal Approaches

Top-performing computer vision models are powered by convolutional neural networks (CNNs). Training an accurate CNN highly depends on both the raw sensor data and their associated ground truth (GT). Collecting such GT is usually done through human labeling, which is time-consuming and does not scale as we wish. This data labeling bottleneck may be intensified due to domain shifts among image sensors, which could force per-sensor data labeling. In this paper, we focus on the use of co-training, a semi-supervised learning (SSL) method, for obtaining self-labeled object bounding boxes (BBs), i.e., the GT to train deep object detectors. In particular, we assess the goodness of multi-modal co-training by relying on two different views of an image, namely, appearance (RGB) and estimated depth (D). Moreover, we compare appearance-based single-modal co-training with multi-modal. Our results suggest that in a standard SSL setting (no domain shift, a few human-labeled data) and under virtual-to-real domain shift (many virtual-world labeled data, no human-labeled data) multi-modal co-training outperforms single-modal. In the latter case, by performing GAN-based domain translation both co-training modalities are on pair; at least, when using an off-the-shelf depth estimation model not specifically trained on the translated images.

cs.CV↗

arXiv2020

Spitzer Observations of the Predicted Eddington Flare from Blazar OJ 287

Binary black hole (BH) central engine description for the unique blazar OJ 287 predicted that the next secondary BH impact-induced bremsstrahlung flare should peak on 2019 July 31. This prediction was based on detailed general relativistic modeling of the secondary BH trajectory around the primary BH and its accretion disk. The expected flare was termed the Eddington flare to commemorate the centennial celebrations of now-famous solar eclipse observations to test general relativity by Sir Arthur Eddington. We analyze the multi-epoch Spitzer observations of the expected flare between 2019 July 31 and 2019 September 6, as well as baseline observations during 2019 February-March. Observed Spitzer flux density variations during the predicted outburst time display a strong similarity with the observed optical pericenter flare from OJ 287 during 2007 September. The predicted flare appears comparable to the 2007 flare after subtracting the expected higher base-level Spitzer flux densities at 3.55 and 4.49 $μ$m compared to the optical R-band. Comparing the 2019 and 2007 outburst lightcurves and the previously calculated predictions, we find that the Eddington flare arrived within 4 hours of the predicted time. Our Spitzer observations are well consistent with the presence of a nano-Hertz gravitational wave emitting spinning massive binary BH that inspirals along a general relativistic eccentric orbit in OJ 287. These multi-epoch Spitzer observations provide a parametric constraint on the celebrated BH no-hair theorem.

astro-ph.HE↗

arXiv2020

Probing the innermost regions of AGN jets and their magnetic fields with RadioAstron. III. Blazar S5 0716+71 at microarcsecond resolution

We present RadioAstron Space VLBI imaging observations of the BL Lac object S5 0716+71 made on January 3-4 2015 at a frequency of 22 GHz (wavelength $λ=1.3$ cm). The observations were made in the framework of the AGN Polarization Key Science Program. The source was detected on projected space-ground baselines up to 70 833 km (5.6 Earth diameters) for both, parallel hand and cross-hand interferometric visibilities. We have used these detections to obtain a full-polarimetric image of the blazar at an unprecedented angular resolution of 24 $μ$as, the highest for this source to date. This enabled us to estimate the size of the radio core to be $<12\times5~μ$as and to reveal a complex structure and a significant curvature of the blazar jet in the inner 100 $μ$as, which is an indication that the jet viewing angle lies inside the opening angle of the jet conical outflow. Fairly highly (15%) linearly polarized emission is detected in a jet region of 19 $μ$as in size, located 58 $μ$as downstream from the core. The highest brightness temperature in the source frame is estimated to be $>2.2\times10^{13}$ K for the blazar core. This implies that the inverse Compton limit must be violated in the rest frame of the source, even for the largest Doppler factor $δ\thicksim25$ reported for 0716+714.

astro-ph.HE↗

arXiv2016

Particle-in-cell Simulations of Global Relativistic Jets with Helical Magnetic Fields

We study the interaction of relativistic jets with their environment, using 3-dimensional relativistic particle-in-cell simulations for two cases of jet composition: (i) electron-proton ($e^{-}-p^{+}$) and (ii) electron-positron ($e^{\pm}$) plasmas containing helical magnetic fields. We have performed simulations of "global" jets containing helical magnetic fields in order to examine how helical magnetic fields affect kinetic instabilities such as the Weibel instability, the kinetic Kelvin-Helmholtz instability and the Mushroom instability. We have found that these kinetic instabilities are suppressed and new types of instabilities can grow. For the $e^{-}-p^{+}$ jet, a recollimation-like instability occurs and jet electrons are strongly perturbed, whereas for the $e^{\pm}$ jet, a recollimation-like instability occurs at early times followed by kinetic instability and the general structure is similar to a simulation without a helical magnetic field. We plan to perform further simulations using much larger systems to confirm these new findings.

astro-ph.HE↗

arXiv2013

High-Angular-Resolution and High-Sensitivity Science Enabled by Beamformed ALMA

An international consortium is presently constructing a beamformer for the Atacama Large Millimeter/submillimeter Array (ALMA) in Chile that will be available as a facility instrument. The beamformer will aggregate the entire collecting area of the array into a single, very large aperture. The extraordinary sensitivity of phased ALMA, combined with the extremely fine angular resolution available on baselines to the Northern Hemisphere, will enable transformational new very long baseline interferometry (VLBI) observations in Bands 6 and 7 (1.3 and 0.8 mm) and provide substantial improvements to existing VLBI arrays in Bands 1 and 3 (7 and 3 mm). The ALMA beamformer will have impact on a variety of scientific topics, including accretion and outflow processes around black holes in active galactic nuclei (AGN), tests of general relativity near black holes, jet launch and collimation from AGN and microquasars, pulsar and magnetar emission processes, the chemical history of the universe and the evolution of fundamental constants across cosmic time, maser science, and astrometry.

astro-ph.IM↗