arXiv subjects

Hong-Seok Kim

Publications and source records attributed to Hong-Seok Kim.

At least 19 recordsLinked to original sources

arXiv2026

Quantum frequency comb with pump-selectable bin pairing and extraction-aware loading in a lithium niobate microresonator

Integrated quantum photonics requires bright, high-fidelity photon-pair sources capable of spectral multiplexing, correlation control, and circuit-compatible extraction. Cavity-enhanced spontaneous parametric down-conversion (SPDC) increases pair generation, but triply resonant operation imposes stringent pump-signal-idler spectral-alignment constraints. Moreover, the trade-off between intrinsic generation and coincidence-to-accidental ratio (CAR) does not capture the usable output flux, which depends on photon extraction. Here, we demonstrate a single-pass-pumped, resonator-enhanced quantum frequency comb (QFC) source based on a periodically poled lithium niobate photonic-crystal Fabry-P$\'e$rot microresonator. The device yields intrinsic and loaded brightnesses of 69.9 and 1.88 MHz/$\mu$W, respectively, and a maximum CAR of 16,000. Frequency-resolved measurements reveal 461 cavity-defined bins spanning 1495-1570 nm, and loaded spectral brightness approaching $4.29\times10^9$ pairs/(s$\cdot$mW$\cdot$nm). Heralded autocorrelation measurements yield a second-order autocorrelation $g_H^{(2)}(0)$ as low as $4.0\times10^{-4}$ for a selected high-CAR cavity-defined bin pair. In particular, tuning the single-pass pump deterministically selects correlated frequency-bin pairings within the fixed QFC grid while preserving brightness and pairwise coincidence rates. We further separate intrinsic generation from output photon-pair flux, revealing the loaded-brightness-CAR relation. Together, pump-selectable bin pairing and extraction-aware loading point to tailored SPDC QFCs as chip-integrated nonclassical light resources for multichannel quantum-state transport and information processing.

physics.optics↗

arXiv2025

High-Q photonic crystal Fabry-Perot micro-resonator in thin-film lithium niobate

Thin-film lithium niobate (TFLN) has emerged as a powerful platform for integrated nonlinear and quantum photonics, owing to its strong optical nonlinearities, wide transparency window, and electro- and piezo-optic properties. However, conventional traveling-wave resonators, such as micro-rings, disks, and racetracks, suffer from curvature-dependent group dispersion and losses, limited spectral tunability, and parasitic nonlinearities, which constrain their performance, scalability, and operational stability in nonlinear photonic circuits. Here, we present photonic crystal (PhC) Fabry-Perot (FP) micro-resonators in TFLN that address these limitations. The device features a one-dimensional straight cavity bounded by PhC reflectors and supports well-confined standing-wave resonant modes within an engineered photonic bandgap. We achieve intrinsic quality (Q) factors of up to 1.4e6 and demonstrate that both the free spectral range (FSR) and coupling strength can be consistently controlled via cavity length and PhC coupler design, respectively. The photonic bandgap is tunable across the S-, C-, and L-bands without degradation of resonator performance. Spectral confinement of high-Q resonant modes is expected to mitigate parasitic nonlinearities, such as Raman scattering. These advances, together with the one-dimensional geometry, establish PhC FP micro-resonators as compact and scalable building blocks for high-density photonic integrated circuits targeting next-generation nonlinear and quantum applications.

physics.optics↗

arXiv2024

MetaMix: Meta-state Precision Searcher for Mixed-precision Activation Quantization

Mixed-precision quantization of efficient networks often suffer from activation instability encountered in the exploration of bit selections. To address this problem, we propose a novel method called MetaMix which consists of bit selection and weight training phases. The bit selection phase iterates two steps, (1) the mixed-precision-aware weight update, and (2) the bit-search training with the fixed mixed-precision-aware weights, both of which combined reduce activation instability in mixed-precision quantization and contribute to fast and high-quality bit selection. The weight training phase exploits the weights and step sizes trained in the bit selection phase and fine-tunes them thereby offering fast training. Our experiments with efficient and hard-to-quantize networks, i.e., MobileNet v2 and v3, and ResNet-18 on ImageNet show that our proposed method pushes the boundary of mixed-precision quantization, in terms of accuracy vs. operations, by outperforming both mixed- and single-precision SOTA methods.

cs.LG↗

arXiv2023

JaxPruner: A concise library for sparsity research

This paper introduces JaxPruner, an open-source JAX-based pruning and sparse training library for machine learning research. JaxPruner aims to accelerate research on sparse neural networks by providing concise implementations of popular pruning and sparse training algorithms with minimal memory and latency overhead. Algorithms implemented in JaxPruner use a common API and work seamlessly with the popular optimization library Optax, which, in turn, enables easy integration with existing JAX based libraries. We demonstrate this ease of integration by providing examples in four different codebases: Scenic, t5x, Dopamine and FedJAX and provide baseline experiments on popular benchmarks.

cs.LG↗

arXiv2023

Video-kMaX: A Simple Unified Approach for Online and Near-Online Video Panoptic Segmentation

Video Panoptic Segmentation (VPS) aims to achieve comprehensive pixel-level scene understanding by segmenting all pixels and associating objects in a video. Current solutions can be categorized into online and near-online approaches. Evolving over the time, each category has its own specialized designs, making it nontrivial to adapt models between different categories. To alleviate the discrepancy, in this work, we propose a unified approach for online and near-online VPS. The meta architecture of the proposed Video-kMaX consists of two components: within clip segmenter (for clip-level segmentation) and cross-clip associater (for association beyond clips). We propose clip-kMaX (clip k-means mask transformer) and HiLA-MB (Hierarchical Location-Aware Memory Buffer) to instantiate the segmenter and associater, respectively. Our general formulation includes the online scenario as a special case by adopting clip length of one. Without bells and whistles, Video-kMaX sets a new state-of-the-art on KITTI-STEP and VIPSeg for video panoptic segmentation, and VSPW for video semantic segmentation. Code will be made publicly available.

cs.CV↗

arXiv2023

TubeFormer-DeepLab: Video Mask Transformer

We present TubeFormer-DeepLab, the first attempt to tackle multiple core video segmentation tasks in a unified manner. Different video segmentation tasks (e.g., video semantic/instance/panoptic segmentation) are usually considered as distinct problems. State-of-the-art models adopted in the separate communities have diverged, and radically different approaches dominate in each task. By contrast, we make a crucial observation that video segmentation tasks could be generally formulated as the problem of assigning different predicted labels to video tubes (where a tube is obtained by linking segmentation masks along the time axis) and the labels may encode different values depending on the target task. The observation motivates us to develop TubeFormer-DeepLab, a simple and effective video mask transformer model that is widely applicable to multiple video segmentation tasks. TubeFormer-DeepLab directly predicts video tubes with task-specific labels (either pure semantic categories, or both semantic categories and instance identities), which not only significantly simplifies video segmentation models, but also advances state-of-the-art results on multiple video segmentation benchmarks

cs.CV↗

arXiv2020

Gate-Modulated Quantum Interference Oscillations in Sb-Doped Bi2Se3 Topological Insulator Nanoribbon

Topological insulator nanoribbons (TI NRs) provide a useful platform to explore the phase-coherent quantum electronic transport of topological surface states, which is crucial for the development of topological quantum devices. When applied with an axial magnetic field, the TI NR exhibits magnetoconductance (MC) oscillations with a flux period of h/e, i.e., Aharonov-Bohm (AB) oscillations, and h/2e, i.e., Altshuler-Aronov-Spivak (AAS) oscillations. Herein, we present an extensive study of the AB and AAS oscillations in Sb doped Bi$_2$Se$_3$ TI NR as a function of the gate voltage, revealing phase-alternating topological AB oscillations. Moreover, the ensemble-averaged fast Fourier transform analysis on the Vg dependent MC curves indicates the suppression of the quantum interference oscillation amplitudes near the Dirac point, which is attributed to the suppression of the phase coherence length within the low carrier density region. The weak antilocalization analysis on the perpendicular MC curves confirms the idea of the suppressed coherence length near the Dirac point in the TI NR.

cond-mat.mes-hall↗

arXiv2020

Adjustable quantum interference oscillations in Sb-doped Bi2Se3 topological insulator nanoribbons

Topological insulator (TI) nanoribbons (NRs) provide a unique platform for investigating quantum interference oscillations combined with topological surface states. One-dimensional subbands formed along the perimeter of a TI NR can be modulated by an axial magnetic field, exhibiting Aharonov-Bohm (AB) and Altshuler-Aronov-Spivak (AAS) oscillations of magnetoconductance (MC). Using Sb-doped Bi2Se3 TI NRs, we found that the relative amplitudes of the two quantum oscillations can be tuned by varying the channel length, exhibiting crossover from quasi-ballistic to diffusive transport regimes. The AB and AAS oscillations were discernible even for a 70 micrometer long channel, while only the AB oscillations were observed for a short channel. Analyses based on ensemble-averaged fast Fourier transform of MC curves revealed exponential temperature dependences of the AB and AAS oscillations, from which the circumferential phase-coherence length and thermal length were obtained. Our observations indicate that the channel length in a TI NR can be a useful control knob for tailored quantum interference oscillations, especially for developing topological hybrid quantum devices.

cond-mat.mes-hall↗

arXiv2020

Superconducting Quantum Interference Devices Made of Sb-doped Bi2Se3 Topological Insulator Nanoribbons

We report the fabrication and characterization of superconducting quantum interference devices (SQUIDs) made of Sb-doped Bi2Se3 topological insulator (TI) nanoribbon (NR) contacted with PbIn superconducting electrodes. When an external magnetic field was applied along the NR axis, the TI NR exhibited periodic magneto-conductance oscillations, the so-called Aharonov-Bohm oscillations, owing to one-dimensional subbands. Below the superconducting transition temperature of PbIn electrodes, we observed supercurrent flow through TI NR-based SQUID. The critical current periodically modulates with a magnetic field perpendicular to the SQUID loop, revealing that the periodicity corresponds to the superconducting flux quantum. Our experimental observations can be useful to explore Majorana bound states (MBS) in TI NR, promising for developing topological quantum information devices.

cond-mat.mes-hall↗

arXiv2019

Towards searching for Majorana fermions in topological insulator nanowires

Developing a gate-tunable, scalable, and topologically-protectable supercurrent qubit and integrating it into a quantum circuit are crucial for applications in the fields of quantum information technology and topological phenomena. Here we propose that the nano-hybrid supercurrent transistors, a superconducting quantum analogue of a transistor, made of topological insulator nanowire would be a promising platform for unprecedented control of both the supercurrent magnitude and the current-phase relation by applying a voltage on a gate electrode. We believe that our experimental design will help probing Majorana state in topological insulator nanowire and establishing a solid-state platform for topological supercurrent qubit.

cond-mat.mes-hall↗

arXiv2019

Electrical detection of spin-polarized current in topological insulator Bi1.5Sb0.5Te1.7Se1.3

Spin-momentum locked (SML) topological surface state (TSS) provides exotic properties for spintronics applications. The spin-polarized current, which emerges owing to the SML, can be directly detected by performing spin potentiometric measurement. We observed spin-polarized current using a bulk insulating topological insulator (TI), Bi1.5Sb0.5Te1.7Se1.3, and Co as the ferromagnetic spin probe. The spin voltage was probed with varying the bias current, temperature, and gate voltage. Moreover, we observed non-local spin-polarized current, which is regarded as a distinguishing property of TIs. The spin-polarization ratio of the non-local current was larger than that of the local current. These findings could reveal a more accurate approach to determine spin-polarization ratio at the TSS.

cond-mat.mes-hall↗

arXiv2019

NNStreamer: Stream Processing Paradigm for Neural Networks, Toward Efficient Development and Execution of On-Device AI Applications

We propose nnstreamer, a software system that handles neural networks as filters of stream pipelines, applying the stream processing paradigm to neural network applications. A new trend with the wide-spread of deep neural network applications is on-device AI; i.e., processing neural networks directly on mobile devices or edge/IoT devices instead of cloud servers. Emerging privacy issues, data transmission costs, and operational costs signifies the need for on-device AI especially when a huge number of devices with real-time data processing are deployed. Nnstreamer efficiently handles neural networks with complex data stream pipelines on devices, improving the overall performance significantly with minimal efforts. Besides, nnstreamer simplifies the neural network pipeline implementations and allows reusing off-shelf multimedia stream filters directly; thus it reduces the developmental costs significantly. Nnstreamer is already being deployed with a product releasing soon and is open source software applicable to a wide range of hardware architectures and software platforms.

cs.DC↗

arXiv2018

Fabrication and Characterization of Superconducting Quantum Interference Device using (Bi_{1-x}Sb_x)_2Se_3 Topological Insulator Nanoribbons

We report on the fabrication and electrical transport properties of superconducting quantum interference devices (SQUIDs) made from a (Bi_{1-x}Sb_x)_2Se_3 topological insulator (TI) nanoribbon (NR) connected with Pb0.5In0.5 superconducting electrodes. Below the transition temperature of the superconducting Pb0.5In0.5 electrodes, periodic oscillations of the critical current are observed in the TI NR SQUID under a magnetic field applied perpendicular to the plane owing to flux quantization. Also the output voltage modulates as a function of the external magnetic field. Moreover, the SQUID the SQUID shows a voltage modulation envelope, which is considered to represent the Fraunhofer-like patterns of each single junction. These properties of the TI NR SQUID would provide a useful method to explore Majorana fermions.

cond-mat.supr-con↗

arXiv2017

Zero bias conductance peak in InAs nanowire coupled to superconducting electrodes

We report the occurrence of the zero-bias conductance peak (ZBCP) in an InAs nanowire coupled to PbIn superconductors with varying temperature, bias voltage, and magnetic field. The ZBCP is suppressed with increasing temperature and bias voltage above the Thouless energy of the nanowire. Applying a magnetic field also diminishes the ZBCP when the resultant magnetic flux reaches the magnetic flux quantum h/2e. Our observations are consistent with theoretical expectations of reflectionless tunneling, in which the phase coherence between an electron and its Andreev-reflected hole induces the ZBCP as long as time-reversal symmetry is preserved.

cond-mat.mes-hall↗

arXiv2017

Macroscopic Quantum Tunneling in Superconducting Junctions of β-Ag$_{2}$Se Topological Insulator Nanowire

We report on the fabrication and electrical transport properties of superconducting junctions made of β-Ag$_{2}$Se topological insulator (TI) nanowires in contact with Al superconducting electrodes. The temperature dependence of the critical current indicates that the superconducting junction belongs to a short and diffusive junction regime. As a characteristic feature of the narrow junction, the critical current decreases monotonously with increasing magnetic field. The stochastic distribution of the switching current exhibits the macroscopic quantum tunneling behavior, which is robust up to T = 0.8 K. Our observations indicate that the TI nanowire-based Josephson junctions can be a promising building block for the development of nanohybrid superconducting quantum bits.

cond-mat.mes-hall↗

arXiv2016

Fabrication and Characterization of PbIn-Au-PbIn Superconducting Junctions

We report on the fabrication and measurement results of the electrical transport properties of superconductor-normal metal-superconductor (SNS) weak links, made of PbIn superconductor and Au metal. The maximum supercurrent reaches up to ~ 6 μA at T = 2.3 K and the supercurrent persists even at higher temperature of T = 4.7 K. Magnetic field dependence of the critical current is consistent with a theoretical fit using the narrow junction model. The superconducting quantum interference device (SQUID) was also fabricated using two PbIn-Au-PbIn junctions connected in parallel. Under perpendicular magnetic field, we clearly observed periodic oscillations of dV/dI with a period of magnetic flux quantum threading into the supercurrent loop of the SQUID. Our fabrication methods would provide an easy and simple way to explore the superconducting proximity effects without ultra-low-temperature cryostats.

cond-mat.supr-con↗

arXiv2016

Very Strong Superconducting Proximity Effects in PbS Semiconductor Nanowires

We report the fabrication of strongly coupled nanohybrid superconducting junctions using PbS semiconductor nanowires and Pb0.5In0.5 superconducting electrodes. The maximum supercurrent in the junction reaches up to ~15 μA at 0.3 K, which is the highest value ever observed in semiconductor-nanowire-based superconducting junctions. The observation of microwave-induced constant voltage steps confirms the existence of genuine Josephson coupling through the nanowire. Monotonic suppression of the critical current under an external magnetic field is also in good agreement with the narrow junction model. The temperature-dependent stochastic distribution of the switching current exhibits a crossover from phase diffusion to a thermal activation process as the temperature decreases. These strongly coupled nanohybrid superconducting junctions would be advantageous to the development of gate-tunable superconducting quantum information devices.

cond-mat.supr-con↗

arXiv2016

Quantum electronic transport of topological surface states in beta-Ag2Se nanowire

Single-crystalline β-Ag2Se nanostructures, a new class of 3D topological insulators (TIs), were synthesized using the chemical vapor transport method. The topological surface states were verified by measuring electronic transport properties including the weak antilocalization effect, Aharonov-Bohm oscillations, and Shubnikov-de Haas oscillations. First-principles band calculations revealed that the band inversion in \b{eta}-Ag2Se is caused by strong spin-orbit coupling and Ag-Se bonding hybridization. These extensive investigations provide new meaningful information about silver-chalcogenide TIs that have anisotropic Dirac cones, which could be useful for spintronics applications.

cond-mat.mes-hall↗