arXiv subjects

Yabin Jin

Publications and source records attributed to Yabin Jin.

9 recordsLinked to original sources

arXiv2026

GA-VINO: A Geometry-Aware Variational Physics-informed Neural Operator for Mindlin-Reissner Plates

Plate and shell structures are widely used in engineering fields. Rapid response prediction for such structures under complex geometries, heterogeneous materials, and varying loads is important for engineering design, but conventional numerical methods usually require repeated modeling and solution when the physical configuration changes. To address this issue, this study proposes a geometry-aware variational physics-informed neural operator (GA-VINO) for Mindlin-Reissner plates. GA-VINO represents the plate geometry using boundary point clouds and incorporates a material encoder, a load encoder, and a scalar-parameter branch to handle spatially random material fields, spatially varying pressure loads, and sample-level uniform parameters. Through multi-branch point cloud encoding and cross-attention, GA-VINO fuses geometric, material, loading, and query point information, and predicts the transverse deflection and rotations at arbitrary query locations. Unlike conventional data-driven neural operators, GA-VINO requires no labeled solution data during training. Instead, it minimizes a variational physics-informed loss constructed from the discretized total potential energy of the Mindlin-Reissner plate. Compared with grid-based neural operators, GA-VINO directly processes irregular point clouds and allows different physical fields to be discretized on different point sets, avoiding forced interpolation onto a common grid. The method is validated on multiple examples involving different geometries, material fields, and load distributions. The results show that GA-VINO achieves promising accuracy in deflection, rotation, gradient-sensitive, and energy-based metrics, completes full-field inference for new samples within milliseconds, and exhibits promising cross-geometry generalization capability.

cs.AI↗

arXiv2026

NeuroSymb-MRG: Differentiable Abductive Reasoning with Active Uncertainty Minimization for Radiology Report Generation

Automatic generation of radiology reports seeks to reduce clinician workload while improving documentation consistency. Existing methods that adopt encoder-decoder or retrieval-augmented pipelines achieve progress in fluency but remain vulnerable to visual-linguistic biases, factual inconsistency, and lack of explicit multi-hop clinical reasoning. We present NeuroSymb-MRG, a unified framework that integrates NeuroSymbolic abductive reasoning with active uncertainty minimization to produce structured, clinically grounded reports. The system maps image features to probabilistic clinical concepts, composes differentiable logic-based reasoning chains, decodes those chains into templated clauses, and refines the textual output via retrieval and constrained language-model editing. An active sampling loop driven by rule-level uncertainty and diversity guides clinician-in-the-loop adjudication and promptbook refinement. Experiments on standard benchmarks demonstrate consistent improvements in factual consistency and standard language metrics compared to representative baselines.

cs.CV↗

arXiv2026

SwiftRepertoire: Few-Shot Immune-Signature Synthesis via Dynamic Kernel Codes

Repertoire-level analysis of T cell receptors offers a biologically grounded signal for disease detection and immune monitoring, yet practical deployment is impeded by label sparsity, cohort heterogeneity, and the computational burden of adapting large encoders to new tasks. We introduce a framework that synthesizes compact task-specific parameterizations from a learned dictionary of prototypes conditioned on lightweight task descriptors derived from repertoire probes and pooled embedding statistics. This synthesis produces small adapter modules applied to a frozen pretrained backbone, enabling immediate adaptation to novel tasks with only a handful of support examples and without full model fine-tuning. The architecture preserves interpretability through motif-aware probes and a calibrated motif discovery pipeline that links predictive decisions to sequence-level signals. Together, these components yield a practical, sample-efficient, and interpretable pathway for translating repertoire-informed models into diverse clinical and research settings where labeled data are scarce and computational resources are constrained.

cs.LG↗

arXiv2021

Inverse design of reconfigurable piezoelectric topological phononic plates

We present a methodology to perform inverse analysis on reconfigurable topological insulators for flexural waves in plate-like structures. First the unit cell topology of a phononic plate is designed, which offers two-fold degeneracy in the band structure by topology optimization. In the second step, piezoelectric patches bonded over the substrate plate are connected to an external circuit and used appropriately to break space inversion symmetry. The space inversion symmetry breaking opens a topological band gap by mimicking quantum valley Hall effect. Numerical simulations demonstrate that the topologically protected edge state exhibits wave propagation without backscattering and is immune to disorders. Predominantly, the proposed idea enables real-time reconfigurability of the topological interfaces in waveguide applications.

cond-mat.mes-hall↗

arXiv2018

Gradient index phononic crystals and metamaterials

Phononic crystals and acoustic metamaterials are periodic structures whose effective properties can be tailored at will to achieve extreme control on wave propagation. Their refractive index is obtained from the homogenization of the infinite periodic system, but it is possible to locally change the properties of a finite crystal in such a way that it results in an effective gradient of the refractive index (GRIN). In such case the propagation of waves can be accurately described by means of ray theory, and different refractive devices can be designed in the framework of wave propagation in inhomogeneous media. In this paper we review the different devices that have been studied for the control of both bulk or guided acoustic waves based on graded phononic crystals.

physics.app-ph↗

arXiv2018

Engineered diffraction gratings for acoustic cloaking

We show that engineered diffraction gratings can considerably simplify the design and performance of acoustic devices. Acoustic reflecting gratings based on rectangular cavities drilled on an acoustically rigid surface are designed in such a way that all the incident energy is channeled towards the diffracted mode traveling in the oposite direction of the incident field (retroreflection effect), and this effect is used to cloak an object placed over a rigid surface. Axisymmetric gratings consisting in rigid surfaces with just one groove per unit cell are used to design acoustic thin acoustic carpet cloaks. Finally, full wave numerical simulations are performed and a conical carpet cloak is experimentally tested, showing an excellent scattering cancellation effect.

physics.app-ph↗

arXiv2018

Analysis of the Robustness of Conventional and Topologically Protected Edge States in Phononic Crystal Plates

In this work we theoretically study the interface acoustic states of resonators on a thin plate with topologically protected and conventional designs. Topologically protected interface state is first analyzed by employing the conception of breaking inversion symmetry within the unit cell of a honeycomb lattice for cylindrical and spherical resonators; we further demonstrate the robustness of the wave propagation along a zig-zag path containing sharp corners, defect and disorder. The wave propagation ceases to be preserved if we increase the degree of disorder along the zig-zag path. In parallel, the conventional interface state is also designed and compared to the same situations. We found that the conventional interface state suffers back scattering in the zig-zag path while it can show a more confined wave transport in some cases. The presence of a defect along the propagation path scatters the conventional interface wave and in particular can prohibit a full propagation in presence of a localized state at the defect. If the zig-zag path is made disordered, the propagation of the conventional interface mode can be conserved at given frequencies for a low random degree and disappears for higher random degree as the interface bands become flat in dispersion and turn to localized states. Finally, we show that the immunity of the topologically protected design needs the interface to be surrounded by at least two hexagons of the phononic crystals on both sides, especially at the sharp corners in the zig-zag path, while the conventional design only needs one hexagon bulk media with the advantage of compact wave transport. This work puts a step forward for the interface states in micro-/nano-scale characterization and figures out the behaviors for both topologically protected and conventional interface states.

cond-mat.mtrl-sci↗

arXiv2015

Gradient Index Devices for the Simultaneous Focusing of the S0 and A0 Lamb Modes

Gradient index devices for the simultaneous control of the S 0 and A 0 Lamb modes are presented and numerically analysed. It is found a special relationship between the refractive indexes for these two modes which allows their simultaneous control by means of graded phononic crystals and thickness variations of the plate. A flat gradient index lens and a circular Luneburg lens are designed. Numerical simulations show that the performance of these devices is good for the two modes in a broadband frequency region, and that this approach can be used to design more advanced refractive devices for the total control of guided vibrational modes.

physics.class-ph↗

arXiv2013

YHap: software for probabilistic assignment of Y haplogroups from population re-sequencing data

Y haplogroup analyses are an important component of genealogical reconstruction, population genetic analyses, medical genetics and forensics. These fields are increasingly moving towards use of low-coverage, high throughput sequencing. However, there is as yet no software available for using sequence data to assign Y haplogroup groups probabilistically, such that the posterior probability of assignment fully reflects the information present in the data, and borrows information across all samples sequenced from a population. YHap addresses this problem.

q-bio.PE↗