arXiv subjects

Robin Wang

Publications and source records attributed to Robin Wang.

5 recordsLinked to original sources

arXiv2026

A multitask framework for automated interpretation of multi-frame right upper quadrant ultrasound in clinical decision support

Ultrasound is a cornerstone of emergency and hepatobiliary imaging, yet its interpretation remains highly operator-dependent and time-sensitive. Here, we present a multitask vision-language agent (VLM) developed to assist with comprehensive right upper quadrant (RUQ) ultrasound interpretation across the full diagnostic workflow. The system was trained on a large, multi-center dataset comprising a primary cohort from Johns Hopkins Medical Institutions (9,189 cases, 594,099 images) and externally validated on cohorts from Stanford University (108 cases, 3,240 images) and a major Chinese medical center (257 cases, 3,178 images). Built on the Qwen2.5-VL-7B architecture, the agent integrates frame-level visual understanding with report-grounded language reasoning to perform three tasks: (i) classification of 18 hepatobiliary and gallbladder conditions, (ii) generation of clinically coherent diagnostic reports, and (iii) surgical decision support based on ultrasound findings and clinical data. The model achieved high diagnostic accuracy across all tasks, generated reports that were indistinguishable from expert-written versions in blinded evaluations, and demonstrated superior factual accuracy and information density on content-based metrics. The agent further identified patients requiring cholecystectomy with high precision, supporting real-time decision-making. These results highlight the potential of generalist vision-language models to improve diagnostic consistency, reporting efficiency, and surgical triage in real-world ultrasound practice.

eess.IV↗

arXiv2025

Navigating the AI-Energy Nexus with Geopolitical Insight

This working paper examines how geopolitical strategies and energy resource management intersect with Artificial Intelligence (AI) development, delineating the AI-energy nexus as critical to sustaining U.S. AI leadership. By analyzing the centralized approaches of authoritarian regimes like China and Gulf nations, alongside market-driven approaches in the U.S., the paper explores divergent strategies to allocate resources for AI energy needs. It underscores the role of energy infrastructure, market dynamics, and state-led initiatives in shaping global AI competition. Recommendations include adopting geopolitically informed analyses and leveraging both market and non-market strengths to enhance U.S. competitiveness. This research aims to inform policymakers, technologists, and researchers about the strategic implications of the AI-energy nexus and offers insights into advancing U.S. global leadership in AI amidst evolving technological paradigms.

cs.CY↗

arXiv2022

ART-Point: Improving Rotation Robustness of Point Cloud Classifiers via Adversarial Rotation

Point cloud classifiers with rotation robustness have been widely discussed in the 3D deep learning community. Most proposed methods either use rotation invariant descriptors as inputs or try to design rotation equivariant networks. However, robust models generated by these methods have limited performance under clean aligned datasets due to modifications on the original classifiers or input space. In this study, for the first time, we show that the rotation robustness of point cloud classifiers can also be acquired via adversarial training with better performance on both rotated and clean datasets. Specifically, our proposed framework named ART-Point regards the rotation of the point cloud as an attack and improves rotation robustness by training the classifier on inputs with Adversarial RoTations. We contribute an axis-wise rotation attack that uses back-propagated gradients of the pre-trained model to effectively find the adversarial rotations. To avoid model over-fitting on adversarial inputs, we construct rotation pools that leverage the transferability of adversarial rotations among samples to increase the diversity of training data. Moreover, we propose a fast one-step optimization to efficiently reach the final robust model. Experiments show that our proposed rotation attack achieves a high success rate and ART-Point can be used on most existing classifiers to improve the rotation robustness while showing better performance on clean datasets than state-of-the-art methods.

cs.CV↗

arXiv2020

Nonlinear wakefields and electron injection in cluster plasma

Laser and beam driven wakefields promise orders of magnitude increases in electric field gradients for particle accelerators for future applications. Key areas to explore include the emittance properties of the generated beams and overcoming the dephasing limit in the plasma. In this paper, the first in-depth study of the self-injection mechanism into wakefield structures from non-homogeneous cluster plasmas is provided using high-resolution two dimensional particle-in-cell simulations. The clusters which are typical structures caused by ejection of gases from a high-pressure gas jet have a diameter much smaller than the laser wavelength. Conclusive evidence is provided for the underlying mechanism that leads to particle trapping, comparing uniform and cluster plasma cases. The accelerated electron beam properties are found to be tunable by changing the cluster parameters. The mechanism explains enhanced beam charge paired with large transverse momentum and energy which has implications for the betatron x-ray flux. Finally, the impact of clusters on the high-power laser propagation behavior is discussed.

physics.plasm-ph↗

arXiv2019

Orbital angular momentum coupling in elastic photon-photon scattering

In this letter, we investigate the effect of orbital angular momentum (OAM) on elastic photon-photon scattering in vacuum for the first time. We define exact solutions to the vacuum electro-magnetic wave equation which carry OAM. Using those, the expected coupling between three initialwaves is derived in the framework of an effective field theory based on the Euler-Heisenberg La-grangian and shows that OAM adds a signature to the generated photons thereby greatly improvingthe signal-to-noise ratio. This forms the basis for a proposed high-power laser experiment utilizingquantum optics techniques to filter the generated photons based on their OAM state

physics.optics↗