Search

The order in which the trajectory is executed is a powerful source of information for recognizers. However, there is still no general approach for recovering the trajectory of complex and long handwriting from static images. Complex specimens can result in multiple pen-downs and in a high number of trajectory crossings yielding agglomerations of pixels (also known as clusters). While the scientific literature describes a wide range of approaches for recovering the writing order in handwriting, these approaches nevertheless lack a common evaluation metric. In this paper, we introduce a new system to estimate the order recovery of thinned static trajectories, which allows to effectively resolve the clusters and select the order of the executed pen-downs. We evaluate how knowing the starting points of the pen-downs affects the quality of the recovered writing. Once the stability and sensitivity of the system is analyzed, we describe a series of experiments with three publicly available databases, showing competitive results in all cases. We expect the proposed system, whose code is made publicly available to the research community, to reduce potential confusion when the order of complex trajectories are recovered, and this will in turn make the trajectories recovered to be viable for further applications, such as velocity estimation.

Building upon findings in computational model of handwriting learning and execution, we introduce the concept of stability to explain the difference between the actual movements performed during multiple execution of the subject's signature, and conjecture that the most stable parts of the signature should play a paramount role in evaluating the similarity between a questioned signature and the reference ones during signature verification. We then introduce the Stability Modulated Dynamic Time Warping algorithm for incorporating the stability regions, i.e. the most similar parts between two signatures, into the distance measure between a pair of signatures computed by the Dynamic Time Warping for signature verification. Experiments were conducted on two datasets largely adopted for performance evaluation. Experimental results show that the proposed algorithm improves the performance of the baseline system and compares favourably with other top performing signature verification systems.

Experiments and numerical simulations of inertial particles in underexpanded jets are performed. The structure of the jet is controlled by varying the nozzle pressure ratio, while the influence of particles on emerging shocks and rarefaction patterns is controlled by varying the particle size and mass loading. Ultra-high-speed schlieren and Lagrangian particle tracking are used to experimentally determine the two-phase flow quantities. Three-dimensional simulations are performed using a high-order, low dissipative discretization of the gas phase while particles are tracked individually in a Lagrangian manner. A simple two-way coupling strategy is proposed to handle interphase exchange in the vicinity of shocks. Velocity statistics of each phase are reported for a wide range of pressure ratios, particle sizes, and volume fractions. The extent to which particles affect the location of the Mach disk are quantified and compared to previous work from the literature. Furthermore, a semi-analytic model is presented based on a one-dimensional Fanno flow that takes into account volume displacement by particles and interphase exchange due to drag and heat transfer. The percent shift in Mach disk is found to scale with the mass loading, nozzle pressure ratio, interphase slip velocity, and inversely with the particle diameter.

New methods for generating synthetic handwriting images for biometric applications have recently been developed. The temporal evolution of handwriting from childhood to adulthood is usually left unexplored in these works. This paper proposes a novel methodology for including temporal evolution in a handwriting synthesizer by means of simplifying the text trajectory plan and handwriting dynamics. This is achieved through a tailored version of the kinematic theory of rapid human movements and the neuromotor inspired handwriting synthesizer. The realism of the proposed method has been evaluated by comparing the temporal evolution of real and synthetic samples both quantitatively and subjectively. The quantitative test is based on a visual perception algorithm that compares the letter variability and the number of strokes in the real and synthetic handwriting produced at different ages. In the subjective test, 30 people are asked to evaluate the perceived realism of the evolution of the synthetic handwriting., Comment: Published in Pattern Recognition

During recent years, there here has been a boom in terms of deep learning use for handwriting analysis and recognition. One main application for handwriting analysis is early detection and diagnosis in the health field. Unfortunately, most real case problems still suffer a scarcity of data, which makes difficult the use of deep learning-based models. To alleviate this problem, some works resort to synthetic data generation. Lately, more works are directed towards guided data synthetic generation, a generation that uses the domain and data knowledge to generate realistic data that can be useful to train deep learning models. In this work, we combine the domain knowledge about the Alzheimer's disease for handwriting and use it for a more guided data generation. Concretely, we have explored the use of in-air movements for synthetic data generation.

A collection of the accepted Findings papers that were presented at the 3rd Machine Learning for Health symposium (ML4H 2023), which was held on December 10, 2023, in New Orleans, Louisiana, USA. ML4H 2023 invited high-quality submissions on relevant problems in a variety of health-related disciplines including healthcare, biomedicine, and public health. Two submission tracks were offered: the archival Proceedings track, and the non-archival Findings track. Proceedings were targeted at mature work with strong technical sophistication and a high impact to health. The Findings track looked for new ideas that could spark insightful discussion, serve as valuable resources for the community, or could enable new collaborations. Submissions to the Proceedings track, if not accepted, were automatically considered for the Findings track. All the manuscripts submitted to ML4H Symposium underwent a double-blind peer-review process.

A collection of the extended abstracts that were presented at the 2nd Machine Learning for Health symposium (ML4H 2022), which was held both virtually and in person on November 28, 2022, in New Orleans, Louisiana, USA. Machine Learning for Health (ML4H) is a longstanding venue for research into machine learning for health, including both theoretical works and applied works. ML4H 2022 featured two submission tracks: a proceedings track, which encompassed full-length submissions of technically mature and rigorous work, and an extended abstract track, which would accept less mature, but innovative research for discussion. All the manuscripts submitted to ML4H Symposium underwent a double-blind peer-review process. Extended abstracts included in this collection describe innovative machine learning research focused on relevant problems in health and biomedicine.

This document describes the general process of setting up, running, and analysing disc galaxy simulations using the freely available program Phantom of RAMSES (PoR). This implements Milgromian Dynamics (MOND) with a patch to the RAMSES grid-based $N$-body and hydrodynamical code that uses adaptive mesh refinement. We discuss the procedure of setting up isolated and interacting disc galaxy initial conditions for PoR, running the simulations, and analysing the results. This manual also concisely documents all previously developed MOND simulation codes and the results obtained with them., Comment: 10 pages. The article has been accepted for publication in the Canadian Journal of Physics. Files mentioned in the article can be found here, https://bitbucket.org/SrikanthTN/bonnpor/src/master/

We present the star cluster catalogs for 17 dwarf and irregular galaxies in the $HST$ Treasury Program "Legacy ExtraGalactic UV Survey" (LEGUS). Cluster identification and photometry in this subsample are similar to that of the entire LEGUS sample, but special methods were developed to provide robust catalogs with accurate fluxes due to low cluster statistics. The colors and ages are largely consistent for two widely used aperture corrections, but a significant fraction of the clusters are more compact than the average training cluster. However, the ensemble luminosity, mass, and age distributions are consistent suggesting that the systematics between the two methods are less than the random errors. When compared with the clusters from previous dwarf galaxy samples, we find that the LEGUS catalogs are more complete and provide more accurate total fluxes. Combining all clusters into a composite dwarf galaxy, we find that the luminosity and mass functions can be described by a power law with the canonical index of $-2$ independent of age and global SFR binning. The age distribution declines as a power law, with an index of $\approx-0.80\pm0.15$, independent of cluster mass and global SFR binning. This decline of clusters is dominated by cluster disruption since the combined star formation histories and integrated-light SFRs are both approximately constant over the last few hundred Myr. Finally, we find little evidence for an upper-mass cutoff ($<2\sigma$) in the composite cluster mass function, and can rule out a truncation mass below $\approx10^{4.5}$M$_{\odot}$ but cannot rule out the existence of a truncation at higher masses., Comment: 25 pages; 28 figures (Accepted to MNRAS)

Recently acquired WFC3 UV (F275W and F336W) imaging mosaics under the Legacy Extragalactic UV Survey (LEGUS) combined with archival ACS data of M51 are used to study the young star cluster (YSC) population of this interacting system. Our newly extracted source catalogue contains 2834 cluster candidates, morphologically classified to be compact and uniform in colour, for which ages, masses and extinction are derived. In this first work we study the main properties of the YSC population of the whole galaxy, considering a mass-limited sample. Both luminosity and mass functions follow a power law shape with slope -2, but at high luminosities and masses a dearth of sources is observed. The analysis of the mass function suggests that it is best fitted by a Schechter function with slope -2 and a truncation mass at $1.00\pm0.12\times10^5\ M_\odot$. Through Monte Carlo simulations we confirm this result and link the shape of the luminosity function to the presence of a truncation in the mass function. A mass limited age function analysis, between 10 and 200 Myr, suggests that the cluster population is undergoing only moderate disruption. We observe little variation in the shape of the mass function at masses above $1\times10^4\ M_\odot$, over this age range. The fraction of star formation happening in the form of bound clusters in M51 is $\sim20\%$ in the age range 10 to 100 Myr and little variation is observed over the whole range from 1 to 200 Myr., Comment: 25 pages, 19 figures, 10 tables. Accepted for publication in MNRAS (14 Sep. 2017)

High speed Schlieren videos were produced highlighting the fluid mechanics found in everyday objects. This video (entry 102369) was submitted as part of the Gallery of Fluid Motion 2013, which is a showcase of fluid dynamics videos., Comment: Both a high-resolution version and a low-resolution version of the submitted video are available for download

Shock wave surfing is investigated experimentally in GALCIT's Mach 4.0 Ludwieg Tube. Shock wave surfing occurs when a secondary free-body follows the bow shock formed by a primary free-body; an example of shock wave surfing occurs during meteorite breakup. The free-bodies in the current investigation are nylon spheres. During each run in the Ludwieg tube a high speed camera is used to capture a series of schlieren images; edge tracking software is used to measure the position of each sphere. Velocity and acceleration are had from processing the position data. The radius ratio and initial orientation of the two spheres are varied in the test matrix. The variation of sphere radius ratio and initial angle between the centers of gravity are shown to have a significant effect on the dynamics of the system. The air flow in each fluid dynamics video is from left to right. The Mach number is 4.0 with total pressure in each run of roughly 200kPa. Each run begins with the spheres tethered and no flow. Flow arrives and the tethers are quickly broken. The spheres are then free-bodies with the dominant force being the pressure distribution around the surface of the sphere, which, in the case of the secondary sphere, is strongly manipulated by the shock wave interaction. Gravity is not a considerable body force because of the short test times in these experiments. The secondary spheres are visibly seen to be affected by the bow shock wave of the primary bodies in the form of quite beautiful trajectories. Computational images were obtained from a simulation using the AMROC code. A fully three-dimensional simulation of two free-flying bodies was performed, assuming inviscid flow, whereby the trajectory of each sphere was calculated according to the instantaneous aerodynamic forces experienced., Comment: This is a submission for the fluid dynamics videos section. There are videos included