Kyuhwan Lee | Searcharxiv

arXiv2026

ATRIA: Adaptive Traceable ECG Reporting with Iterative Agents

Existing ECG report generation is tightly coupled -- interpretation and reporting fused end-to-end, so errors propagate without stage-level recourse -- while agent-based systems decouple tasks but remain single-pass, never revisiting earlier outputs. Clinical ECG reporting instead unfolds iteratively, requiring progressive context integration and bidirectional editing. We present \textsc{ATRIA}, a multi-agent ECG reporting system that mirrors the clinician's iterative workflow: it binds every report claim to its supporting evidence, flags statements unsupported by that evidence, incorporates additional context mid-session, and lets clinicians verify and revise individual findings rather than accept one opaque output. Because its agents use ECG analysis models already in clinical use, the underlying findings are clinically trustworthy; and as a cloud-based web service, \textsc{ATRIA} is ready for immediate deployment. We demonstrate \textsc{ATRIA} through four interaction cases, with a live demo and video available.

cs.AI↗

arXiv2025

Dual-species atomic absorption image reconstruction using deep neural networks

Optical imaging plays an instrumental role in understanding the behavior of trapped neutral atoms. In this work, we describe a deep learning-based online image completion protocol that reduces interference fringes in optical absorption signals for a dual-species atomic system. Regardless of the distinct nature of the task for two different atomic species, 6Li and 23Na, the method displays a robust solution for suppressing fringes. To incorporate this into daily operations, a transfer learning scheme is required that incrementally updates the previously learned parameters. We outline an online image completion method that efficiently adapts to drifting experimental conditions. Our method can be easily integrated into lab settings, where transfer learning can accelerate image analysis.

cond-mat.quant-gas↗

arXiv2025

On-device Sora: Enabling Training-Free Diffusion-based Text-to-Video Generation for Mobile Devices

We present On-device Sora, the first model training-free solution for diffusion-based on-device text-to-video generation that operates efficiently on smartphone-grade devices. To address the challenges of diffusion-based text-to-video generation on computation- and memory-limited mobile devices, the proposed On-device Sora applies three novel techniques to pre-trained video generative models. First, Linear Proportional Leap (LPL) reduces the excessive denoising steps required in video diffusion through an efficient leap-based approach. Second, Temporal Dimension Token Merging (TDTM) minimizes intensive token-processing computation in attention layers by merging consecutive tokens along the temporal dimension. Third, Concurrent Inference with Dynamic Loading (CI-DL) dynamically partitions large models into smaller blocks and loads them into memory for concurrent model inference, effectively addressing the challenges of limited device memory. We implement On-device Sora on the iPhone 15 Pro, and the experimental evaluations show that it is capable of generating high-quality videos on the device, comparable to those produced by high-end GPUs. These results show that On-device Sora enables efficient and high-quality video generation on resource-constrained mobile devices. We envision the proposed On-device Sora as a significant first step toward democratizing state-of-the-art generative technologies, enabling video generation on commodity mobile and embedded devices without resource-intensive re-training for model optimization (compression). The code implementation is available at a GitHub repository(https://github.com/eai-lab/On-device-Sora).

cs.CV↗

arXiv2025

On-device Sora: Enabling Training-Free Diffusion-based Text-to-Video Generation for Mobile Devices

We present On-device Sora, the first model training-free solution for diffusion-based on-device text-to-video generation that operates efficiently on smartphone-grade devices. To address the challenges of diffusion-based text-to-video generation on computation- and memory-limited mobile devices, the proposed On-device Sora applies three novel techniques to pre-trained video generative models. First, Linear Proportional Leap (LPL) reduces the excessive denoising steps required in video diffusion through an efficient leap-based approach. Second, Temporal Dimension Token Merging (TDTM) minimizes intensive token-processing computation in attention layers by merging consecutive tokens along the temporal dimension. Third, Concurrent Inference with Dynamic Loading (CI-DL) dynamically partitions large models into smaller blocks and loads them into memory for concurrent model inference, effectively addressing the challenges of limited device memory. We implement On-device Sora on the iPhone 15 Pro, and the experimental evaluations show that it is capable of generating high-quality videos on the device, comparable to those produced by high-end GPUs. These results show that On-device Sora enables efficient and high-quality video generation on resource-constrained mobile devices. We envision the proposed On-device Sora as a significant first step toward democratizing state-of-the-art generative technologies, enabling video generation on commodity mobile and embedded devices without resource-intensive re-training for model optimization (compression). The code implementation is available at a GitHub repository(https://github.com/eai-lab/On-device-Sora).

cs.CV↗

arXiv2024

Flattening a trapped atomic gas using a programmable optical potential in a feedback loop

We present a method for producing a flat, large-area Fermi gas of $^6$Li with a uniform area density. The method uses a programmable optical potential within a feedback loop to flatten the in-plane trapping potential for atoms. The optical potential is generated using a laser beam, whose intensity profile is adjusted by a spatial light modulator and optimized through measurements of the density distribution of the sample. The resulting planar sample exhibits a uniform area density within a region of about 480 $μ$m in diameter and the standard deviation of the trap bottom potential is estimated to be $\approx k_B \times$ 6.1 nK, which is less than 20$\%$ of the transverse confinement energy. We discuss a dimensional crossover toward 2D regime by reducing the number of atoms in the planar trap, including the effect of the spatial variation of the transverse trapping frequency in the large-area sample.

cond-mat.quant-gas↗

arXiv2023

Observation of universal Kibble-Zurek scaling in an atomic Fermi superfluid

Half a century ago, T. Kibble proposed a scenario for topological defect formation from symmetry breaking during the expansion of the early Universe. W. Zurek later crystallized the concept to superfluid helium, predicting a power-law relation between the number of quantum vortices and the rate at which the system passes through the lambda transition. Here, we report the observation of Kibble-Zurek scaling in a homogeneous, strongly interacting Fermi gas undergoing a superfluid phase transition. We investigate the superfluid transition using two distinct control parameters: temperature and interaction strength. The microscopic physics of condensate formation is markedly different for the two quench parameters, signaled by their two orders of magnitude difference in the condensate formation timescale. However, regardless of the thermodynamic direction in which the system passes through a phase transition, the Kibble-Zurek exponent is identically observed to be about 0.68 and shows good agreement with theoretical predictions that describe superfluid phase transitions. This work demonstrates the gedanken experiment Zurek proposed for liquid helium that shares the same universality class with strongly interacting Fermi gases.

cond-mat.quant-gas↗

arXiv2023

Spin-driven stationary turbulence in spinor Bose-Einstein condensates

We report the observation of stationary turbulence in antiferromagnetic spin-1 Bose-Einstein condensates driven by a radio-frequency magnetic field. The magnetic driving injects energy into the system by spin rotation and the energy is dissipated via dynamic instability, resulting in the emergence of an irregular spin texture in the condensate. Under continuous driving, the spinor condensate evolves into a nonequilibrium steady state with characteristic spin turbulence, while the low energy scale of spin excitations ensures that the sample's lifetime is minimally affected. When the driving strength is on par with the system's spin interaction energy and the quadratic Zeeman energy, remarkably, the stationary turbulent state exhibits spin-isotropic features in spin composition and spatial spin texture. We numerically show that ambient field fluctuations play a crucial role in sustaining the turbulent state within the system. These results open up new avenues for exploring quantum turbulence in spinor superfluid systems.

cond-mat.quant-gas↗

arXiv2021

Critical Energy Dissipation in a Binary Superfluid Gas by a Moving Magnetic Obstacle

We study the critical energy dissipation in an atomic superfluid gas with two symmetric spin components by an oscillating magnetic obstacle. Above a certain critical oscillation frequency, spin-wave excitations are generated by the magnetic obstacle, demonstrating the spin superfluid behavior of the system. When the obstacle is strong enough to cause density perturbations via local saturation of spin polarization, half-quantum vortices (HQVs) are created for higher oscillation frequencies, which reveals the characteristic evolution of critical dissipative dynamics from spin-wave emission to HQV shedding. Critical HQV shedding is further investigated using a pulsed linear motion of the obstacle, and we identify two critical velocities to create HQVs with different core magnetization.

cond-mat.quant-gas↗