arXiv subjects

Seungjin Lee

Publications and source records attributed to Seungjin Lee.

At least 19 recordsLinked to original sources

arXiv2026

The Thick Part of the $\mathrm{PSL}_n(\mathbb{R})$-Hitchin-Riemann Moduli Space has Infinite Volume

We prove that the thick part of the $\mathrm{PSL}_n(\mathbb{R})$-Hitchin-Riemann moduli space has infinite total Atiyah--Bott--Goldman volume for $n>2$. This result stands in contrast to Mumford's compactness criterion. To achieve this result, we employ Goldman flows and internal sequences to find an infinite series of subsets of identical volume, the images of which in the Hitchin-Riemann moduli space are all mutually disjoint and sit in the thick part.

math.GT↗

arXiv2026

Pseudo-spectral frequency-domain method with background field decomposition and Green's function preconditioner for electromagnetic scattering problem in EUV lithography

We provide an accelerated computational framework to solve electromagnetic scattering problems in planarly layered media arising from extreme ultraviolet (EUV) lithography. To achieve this, we reformulate the EUV scattering problem into a scattering problem on a homogeneous background, in which the electromagnetic contribution of the layered media is captured by a recursively updated reflection of the layered stack. The system is numerically solved by employing the pseudo-spectral frequency-domain method paired with an iterative solver, whose iterative convergence is expedited by a free-space Green's function preconditioner. The proposed framework is evaluated on EUV mask geometries and multilayer mirror stacks, demonstrating a significant speedup over the conventional pseudo-spectral frequency-domain method.

physics.optics↗

arXiv2025

Universal Fabrication of Graphene/Perovskite Oxide Hybrid Heterostructures

Hybrid heterostructures composed of graphene and perovskite oxides provide a promising platform for exploiting synergetic interfacial functionalities. Conventional fabrication methods of the hybrid heterostructures rely on transferring graphene grown on metallic substrates-- a process that is time-consuming, labor-intensive, and prone to introducing numerous defects. In this study, we present a universal, catalyst-free method for the direct growth of graphene on insulating substrates by employing three different perovskite oxide substrates (SrTiO$_3$, LaAlO$_3$, and (La$_{0.18}$Sr$_{0.82}$)(Al$_{0.59}$Ta$_{0.41}$)O$_3$) using atmospheric chemical vapor deposition. Comprehensive characterization via Raman spectroscopy, X-ray spectroscopy, scanning probe microscopy, and electron microscopy confirmed the formation of a uniform, continuous monolayer graphene on all substrates. We identified that growth temperature critically governs graphene quality, as excessive active species may lead to secondary nucleation and the formation of multilayer graphene. Notably, all substrates shared the same optimal growth conditions. Low-temperature Raman spectroscopy and scanning tunneling microscopy of the graphene/SrTiO$_3$ hybrid heterostructure revealed cooperative phenomena, including substrate-induced lattice-phonon and electron-phonon coupling. Our work establishes a reproducible, transfer-free fabrication route for graphene/perovskite oxide hybrid heterostructures and provides empirical support for the universal growth of graphene on insulating substrates.

cond-mat.mtrl-sci↗

arXiv2025

Quantum Amplitude-Amplification Eigensolver: A State-Learning-Assisted Approach beyond Energy-Gradient-Based Heuristics

Ground-state estimation lies at the heart of a broad range of quantum simulations. Most near-term approaches are cast as variational energy minimization and thus inherit the challenges of problem-specific energy landscapes. We develop the quantum amplitude-amplification eigensolver (QAAE), which departs from the variational paradigm and instead coherently drives a trial state toward the ground state via quantum amplitude amplification. Each amplitude-amplification round interleaves a reflection about the learned trial state with a controlled short-time evolution under a normalized Hamiltonian; an ancilla readout yields an amplitude-amplified pure target state that a state-learning step then re-encodes into an ansatz circuit for the next round -- without evaluating the energy gradients. Under standard assumptions (normalized $\hat{H}$, a nondegenerate ground-state, and a learning update), the ground-state overlap increases monotonically per round and the procedure converges; here, a per-round depth bound in terms of the ansatz depth and Hamiltonian-simulation cost establishes hardware compatibility. Cloud experiments on IBMQ processor verify our amplification mechanism on a two-level Hamiltonian and a two-qubit Ising model, and numerical benchmarks on $\mathrm{H}_2$, $\mathrm{LiH}$, and a $10$-qubit longitudinal-and-transverse-field Ising model show that QAAE integrates with chemistry-inspired and hardware-efficient circuits and can surpass gradient-based VQE in accuracy and stability. These results position QAAE as a variational-free and hardware-compatible route to ground-state estimation for near-term quantum simulation.

quant-ph↗

arXiv2025

Beyond Hard Sharing: Efficient Multi-Task Speech-to-Text Modeling with Supervised Mixture of Experts

Hard-parameter sharing is a common strategy to train a single model jointly across diverse tasks. However, this often leads to task interference, impeding overall model performance. To address the issue, we propose a simple yet effective Supervised Mixture of Experts (S-MoE). Unlike traditional Mixture of Experts models, S-MoE eliminates the need for training gating functions by utilizing special guiding tokens to route each task to its designated expert. By assigning each task to a separate feedforward network, S-MoE overcomes the limitations of hard-parameter sharing. We further apply S-MoE to a speech-to-text model, enabling the model to process mixed-bandwidth input while jointly performing automatic speech recognition (ASR) and speech translation (ST). Experimental results demonstrate the effectiveness of the proposed S-MoE, achieving a 6.35% relative improvement in Word Error Rate (WER) when applied to both the encoder and decoder.

cs.CL↗

arXiv2024

Text-Aware Adapter for Few-Shot Keyword Spotting

Recent advances in flexible keyword spotting (KWS) with text enrollment allow users to personalize keywords without uttering them during enrollment. However, there is still room for improvement in target keyword performance. In this work, we propose a novel few-shot transfer learning method, called text-aware adapter (TA-adapter), designed to enhance a pre-trained flexible KWS model for specific keywords with limited speech samples. To adapt the acoustic encoder, we leverage a jointly pre-trained text encoder to generate a text embedding that acts as a representative vector for the keyword. By fine-tuning only a small portion of the network while keeping the core components' weights intact, the TA-adapter proves highly efficient for few-shot KWS, enabling a seamless return to the original pre-trained model. In our experiments, the TA-adapter demonstrated significant performance improvements across 35 distinct keywords from the Google Speech Commands V2 dataset, with only a 0.14% increase in the total number of parameters.

eess.AS↗

arXiv2024

Quantum electrodynamical formulation of photochemical acid generation and its implications on optical lithography

The photochemical acid generation is refined from the first principles of quantum electrodynamics. First, we briefly review the formulation of the quantum theory of light based on the quantum electrodynamics framework to establish the probability of acid generation at a given spacetime point. The quantum-mechanical acid generation is then combined with the deprotection mechanism to obtain a probabilistic description of the deprotection density directly related to feature formation in a photoresist. A statistical analysis of the random deprotection density is presented to reveal the leading characteristics of stochastic feature formation.

physics.chem-ph↗

arXiv2024

Telepathic Datacenters: Fast RPCs using Shared CXL Memory

Datacenter applications often rely on remote procedure calls (RPCs) for fast, efficient, and secure communication. However, RPCs are slow, inefficient, and hard to use as they require expensive serialization and compression to communicate over a packetized serial network link. Compute Express Link 3.0 (CXL) offers an alternative solution, allowing applications to share data using a cache-coherent, shared-memory interface across clusters of machines. RPCool is a new framework that exploits CXL's shared memory capabilities. RPCool avoids serialization by passing pointers to data structures in shared memory. While avoiding serialization is useful, directly sharing pointer-rich data eliminates the isolation that copying data over traditional networks provides, leaving the receiver vulnerable to invalid pointers and concurrent updates to shared data by the sender. RPCool restores this safety with careful and efficient management of memory permissions. Another significant challenge with CXL shared memory capabilities is that they are unlikely to scale to an entire datacenter. RPCool addresses this by falling back to RDMA-based communication. Overall, RPCool reduces the round-trip latency by 1.93$\times$ and 7.2$\times$ compared to state-of-the-art RDMA and CXL-based RPC mechanisms, respectively. Moreover, RPCool performs either comparably or better than other RPC mechanisms across a range of workloads.

cs.DC↗

arXiv2024

Global optimization in variational quantum algorithms via dynamic tunneling method

We present a global optimization routine for the variational quantum algorithms, which utilizes the dynamic tunneling flow. Originally designed to leverage information gathered by a gradient-based optimizer around local minima, we adapt the conventional dynamic tunneling flow to exploit the distance measure of quantum states, resolving issues of extrinsic degeneracy arising from the parametrization of quantum states. Our global optimization algorithm is applied to the variational quantum eigensolver for the transverse-field Ising model to demonstrate the performance of our routine while comparing it with the conventional dynamic tunneling method, which is based on the Euclidean distance measure on the parameter space.

quant-ph↗

arXiv2024

MaLa-ASR: Multimedia-Assisted LLM-Based ASR

As more and more information-rich data like video become available, utilizing multi-modal auxiliary information to enhance audio tasks has sparked widespread research interest. The recent surge in research on LLM-based audio models provides fresh perspectives for tackling audio tasks. Given that LLM can flexibly ingest multiple inputs, we propose MaLa-ASR, an LLM-based ASR model that can integrate textual keywords extracted from presentation slides to improve recognition of conference content. MaLa-ASR yields average WERs of 9.4% and 11.7% on the L95 and S95 subsets of the SlideSpeech corpus, representing a significant relative WER drop of 27.9% and 44.7% over the baseline model reported in SlideSpeech. MaLa-ASR underscores LLM's strong performance in speech tasks and the capability to integrate auxiliary information conveniently. By adding keywords to the input prompt, the biased word error rate (B-WER) reduces relatively by 46.0% and 44.2%, establishing a new SOTA on this dataset.

eess.AS↗

arXiv2024

Relational Proxy Loss for Audio-Text based Keyword Spotting

In recent years, there has been an increasing focus on user convenience, leading to increased interest in text-based keyword enrollment systems for keyword spotting (KWS). Since the system utilizes text input during the enrollment phase and audio input during actual usage, we call this task audio-text based KWS. To enable this task, both acoustic and text encoders are typically trained using deep metric learning loss functions, such as triplet- and proxy-based losses. This study aims to improve existing methods by leveraging the structural relations within acoustic embeddings and within text embeddings. Unlike previous studies that only compare acoustic and text embeddings on a point-to-point basis, our approach focuses on the relational structures within the embedding space by introducing the concept of Relational Proxy Loss (RPL). By incorporating RPL, we demonstrated improved performance on the Wall Street Journal (WSJ) corpus.

eess.AS↗

arXiv2023

Generalized Euler angles for a unitary control of the Hamiltonian system

We provide an angular parametrization of the special unitary group $\textrm{SU}(2^{n})$ generalizing Euler angles for $\textrm{SU}(2)$ by successively applying the KAK decomposition. We then determine constraint equations for the parametric curve of generalized Euler angles corresponding to the exponential curve of a given Hamiltonian. The constraint equations are in the form of first-order differential-algebraic equations and resemble Wei-Norman equations of canonical coordinates of the second kind for $\textrm{SU}(2^{n})$.

quant-ph↗

arXiv2022

Identifying Critical LMS Features for Predicting At-risk Students

Learning management systems (LMSs) have become essential in higher education and play an important role in helping educational institutions to promote student success. Traditionally, LMSs have been used by postsecondary institutions in administration, reporting, and delivery of educational content. In this paper, we present an additional use of LMS by using its data logs to perform data-analytics and identify academically at-risk students. The data-driven insights would allow educational institutions and educators to develop and implement pedagogical interventions targeting academically at-risk students. We used anonymized data logs created by Brightspace LMS during fall 2019, spring 2020, and fall 2020 semesters at our college. Supervised machine learning algorithms were used to predict the final course performance of students, and several algorithms were found to perform well with accuracy above 90%. SHAP value method was used to assess the relative importance of features used in the predictive models. Unsupervised learning was also used to group students into different clusters based on the similarities in their interaction/involvement with LMS. In both of supervised and unsupervised learning, we identified two most-important features (Number_Of_Assignment_Submissions and Content_Completed). More importantly, our study lays a foundation and provides a framework for developing a real-time data analytics metric that may be incorporated into a LMS.

cs.CY↗

arXiv2021

A Policy-based Versioning SSD with Intel SGX

Privileged malware neutralizes software-based versioning systems and destroys data. To counter this threat, a versioning solid-state drive (SSD) that performs versioning inside the SSD has been studied. An SSD is a suitable candidate for data versioning because it can preserve previous versions without additional copying, and provide high security with a very small trusted computing base (TCB). However, the versioning SSDs studied so far commonly use a full disk versioning method that preserves all file versions in a batch. This paper demonstrates that SSDs, which provide full disk versioning, can be exposed to data tampering attacks when the retention time of data is less than the malware's dwell time. To deal with this threat, we propose SGX-SSD, a policy-based per-file versioning SSD to keep a deeper history for only the important files of users. However, since the SSD isn't aware of a file semantic, and the versioning policy information should be securely received from the untrusted host computer, implementing the per-file versioning in SSD is a huge challenge. To solve this problem, SGX-SSD utilizes the Intel SGX and has a secure host interface to securely receive policy information (configuration values) from the user. Also, to solve the file semantic unawareness problem of the SSD, a piggyback module is designed to give a file hint at the host layer, and an algorithm for selective versioning based on the policy is implemented in the SSD. To prove our system, we prototyped SGX-SSD the Jasmine OpenSSD platform in Linux environment. In the experimental evaluation, we proved that SGX-SSD provides strong security with little additional overhead for selective per-file versioning.

cs.CR↗

arXiv2020

Sequential Skip Prediction with Few-shot in Streamed Music Contents

This paper provides an outline of the algorithms submitted for the WSDM Cup 2019 Spotify Sequential Skip Prediction Challenge (team name: mimbres). In the challenge, complete information including acoustic features and user interaction logs for the first half of a listening session is provided. Our goal is to predict whether the individual tracks in the second half of the session will be skipped or not, only given acoustic features. We proposed two different kinds of algorithms that were based on metric learning and sequence learning. The experimental results showed that the sequence learning approach performed significantly better than the metric learning approach. Moreover, we conducted additional experiments to find that significant performance gain can be achieved using complete user log information.

cs.IR↗

arXiv2020

SGX-SSD: A Policy-based Versioning SSD with Intel SGX

This paper demonstrates that SSDs, which perform device-level versioning, can be exposed to data tampering attacks when the retention time of data is less than the malware's dwell time. To deal with that threat, we propose SGX-SSD, a SGX-based versioning SSD which selectively preserves file history based on the given policy. The proposed system adopts Intel SGX to implement the version policy management system that is safe from high-privileged malware. Based on the policy, only the necessary data is selectively preserved in SSD that prevents files with less priority from wasting space and also ensures the integrity of important files.

cs.CR↗

arXiv2018

Content-based feature exploration for transparent music recommendation using self-attentive genre classification

Interpretation of retrieved results is an important issue in music recommender systems, particularly from a user perspective. In this study, we investigate the methods for providing interpretability of content features using self-attention. We extract lyric features with the self-attentive genre classification model trained on 140,000 tracks of lyrics. Likewise, we extract acoustic features using the acoustic model with self-attention trained on 120,000 tracks of acoustic signals. The experimental results show that the proposed methods provide the characteristics that are interpretable in terms of both lyrical and musical contents. We demonstrate this by visualizing the attention weights, and by presenting the most similar songs found using lyric or audio features.

cs.IR↗

arXiv2018

Fermionic one-loop amplitudes of the RNS superstring

We investigate massless n-point one-loop amplitudes of the open RNS superstring with two external fermions and determine their worldsheet integrands. The contributing correlation functions involving spin-1/2 and spin-3/2 operators from the fermion vertices are evaluated to any multiplicity. Moreover, we introduce techniques to sum these correlators over the spin structures of the worldsheet fermions such as to manifest all cancellations due to spacetime supersymmetry. These spin sums require generalizations of the Riemann identities among Jacobi theta functions, and the results can be expressed in terms of doubly-periodic functions known from the mathematics literature on elliptic multiple zeta values. On the boundary of moduli space, our spin-summed correlators specialize to compact representations of fermionic one-loop integrands for ambitwistor strings.

hep-th↗