arXiv subjects

Youngmin Kim

Publications and source records attributed to Youngmin Kim.

At least 19 recordsLinked to original sources

arXiv2026

SpiderLS: Leveraging Full ZX Reduction for Lattice Surgery Compilation

Lattice surgery compilation plays a central role in translating fault-tolerant quantum programs into efficient surface code realizations, where both spatial and temporal resources directly determine the cost of execution. Recent work has demonstrated the benefits of using ZX-diagrams as an intermediate representation for lattice surgery compilation, enabling semantics-preserving transformations that reduce spacetime cost. However, existing compilation restricts ZX reduction to preserve diagram structures that can be directly embedded as lattice surgery junctions. We present SpiderLS, which extends prior approach by leveraging full ZX reduction. To translate the resulting diagram into executable lattice surgery operations, SpiderLS applies a sequence of compiler passes that derives an execution order, generates target code by grouping compatible interactions into multi-target operations, and lowers the target code to Pauli-product measurements. The resulting explicit patch and Pauli-boundary requirements guide logical scheduling and structure-aware spacetime routing. Across representative algorithmic and random workloads, SpiderLS achieves average reductions of 49.2% in spacetime volume and 99.8% in compilation time compared with the prior ZX-based compiler.

quant-ph↗

arXiv2026

DualView: Preventing Indirect Prompt Injection in Personal AI Agents

Personal AI agents that run on the user's local machine, such as OpenClaw, automate daily tasks including web search, email, and file management. Their access to computer resources, including the network, file system, and shell, exposes them to indirect prompt injection (IPI) attacks. Prior Dual LLM defenses block IPI by replacing untrusted data with symbols that the agent can reference but not read. However, they track untrusted data only inside the agent's context, so when the agent saves and later rereads untrusted data, that data, possibly an attacker's prompt, can return as trusted data rather than as a symbol, which we call stored IPI. Operating on the user's real environment, which humans and programs share, is what makes agents like OpenClaw practical, and is exactly why a defense that ignores it is incomplete. Preserving symbols in such an environment is hard, because humans and programs need original data. We present DualView, which extends untrusted data tracking from the agent's context to the user's environment, including the file system, shell, network, and other agents, by giving each channel two views. In AgentView, the agent sees untrusted data as symbols even after writing it out and reading it back, blocking stored IPI, while HumanView preserves original data for humans and tools. DualView routes each tool call to the right view and synchronizes data across the two views. DualView deploys as an OpenClaw plugin using only tool hooks, without changing the agent's tool-call logic or tool implementations. Since DualView isolates untrusted data by design, its protection is not limited to known attack templates. In our evaluation on an IPI benchmark and PinchBench, DualView blocked every IPI attack, including stored IPI, while keeping utility close to the unprotected baseline.

cs.CR↗

arXiv2026

QWERTY: Training-Free Motion Control via Query-Warped Video Diffusion Transformers

Video diffusion transformers (DiTs) generate high-fidelity and temporally coherent videos, yet motion control remains implicit, primarily relying on text prompts. As a result, achieving desired motion often requires extensive prompt engineering and repeated resampling. While fine-tuning models with additional spatial prompts (e.g., bounding boxes or point trajectories) enables explicit control, it demands substantial data curation and computation, and may compromise the generative capabilities of pretrained models. Consequently, training-free motion control using such spatial prompts has been explored in U-Net-based video diffusion models, but remains largely unexplored for DiTs. We introduce QWERTY, a training-free framework that enables flexible motion control in pretrained image-to-video DiTs via user-defined object warping and optical flow. We carefully manipulate the 3D full attention of DiTs by warping the frame-invariant semantic subspace of queries. We find that the noise predicted by the query-warped DiT naturally guides the diffusion trajectory toward the desired motion, and further show that leveraging this noise as self-guidance for latent optimization improves control stability and visual quality. Experiments show that QWERTY achieves the most effective motion control among existing training-free approaches on a recent image-to-video DiT, with performance comparable to fine-tuning-based methods.

cs.CV↗

arXiv2026

Towards Continuous Sign Language Conversation from Isolated Signs

Sign language is the primary language for many Deaf and Hard-of-Hearing (DHH) signers, yet most conversational AI systems still mediate interaction through spoken or written language. This spoken-language-centered interface can limit access for signers for whom spoken or written language is not the most accessible medium, motivating direct sign-to-sign conversational modeling. However, sentence-level sign video data are expensive to collect and annotate, leaving existing sign translation and production models with limited vocabulary coverage and weak open-domain generalization. We address this bottleneck by constructing continuous sign conversations from isolated signs: large-scale labeled isolated clips are collected as lexically grounded motion primitives and recomposed into sign-language-ordered utterances derived from existing dialogue corpora. We introduce SignaVox-W, which provides, to our knowledge, the largest labeled isolated-sign vocabulary to date, and SignaVox-U, a continuous 3D sign conversation dataset built from SignaVox-W. To bridge structural mismatch between spoken and signed languages, we use a retrieval-guided spoken-to-gloss translator; to bridge independently collected isolated clips, we propose BRAID, a diffusion Transformer that performs duration alignment and co-articulatory boundary inpainting. With the resulting data, we train SignaVox, a direct sign-to-sign conversational model that generates 3D body, hand, and facial motion responses from prior signing context without spoken-language text or externally provided glosses at inference time. Quantitative and qualitative evaluations show improved isolated-to-continuous motion quality, stronger response-level semantic alignment, and scalable signer-centered interaction that better supports visual-spatial articulation.

cs.CV↗

arXiv2026

CRePE: Curved Ray Expectation Positional Encoding for Unified-Camera-Controlled Video Generation

Video world models should predict future appearance in a way that remains consistent with 3D scene structure, camera motion, and lens geometry. Existing attention-level camera encodings, however, either describe each token only by its viewing ray---without locating scene content along that ray---or assume pinhole projection, limiting camera control under wide-angle and fisheye lenses. We introduce Curved Ray Expectation Positional Encoding (CRePE), which represents each image token as a depth-aware distribution along its Unified Camera Model (UCM) ray and integrates the expected rotary positional phasor along the curved path this distribution traces when projected into each query view. CRePE is realized through a lightweight Geometric Attention Adapter on a frozen video diffusion transformer, with pseudo radial-distance supervision from a monocular geometry foundation model serving as a stabilizing anchor rather than an inference-time input. CRePE improves camera-control, lens, and orientation fidelity across pinhole, wide-angle, and fisheye settings, and transfers zero-shot to unseen real fisheye and diverse pinhole videos. Through Radial MixForcing, the same positional pathway further accepts externally supplied radial maps, enabling scene-geometry-conditioned generation and source-video motion transfer that follow the supplied geometry more faithfully than dedicated depth-conditioned baselines. CRePE thus offers a compact interface that unifies camera control, implicit 3D scene state, and external geometry control for video world models.

cs.CV↗

arXiv2026

UCMNet: Uncertainty-Aware Context Memory Network for Under-Display Camera Image Restoration

Under-display cameras (UDCs) allow for full-screen designs by positioning the imaging sensor underneath the display. Nonetheless, light diffraction and scattering through the various display layers result in spatially varying and complex degradations, which significantly reduce high-frequency details. Current PSF-based physical modeling techniques and frequency-separation networks are effective at reconstructing low-frequency structures and maintaining overall color consistency. However, they still face challenges in recovering fine details when dealing with complex, spatially varying degradation. To solve this problem, we propose a lightweight \textbf{U}ncertainty-aware \textbf{C}ontext-\textbf{M}emory \textbf{Network} (\textbf{UCMNet}), for UDC image restoration. Unlike previous methods that apply uniform restoration, UCMNet performs uncertainty-aware adaptive processing to restore high-frequency details in regions with varying degradations. The estimated uncertainty maps, learned through an uncertainty-driven loss, quantify spatial uncertainty induced by diffraction and scattering, and guide the Memory Bank to retrieve region-adaptive context from the Context Bank. This process enables effective modeling of the non-uniform degradation characteristics inherent to UDC imaging. Leveraging this uncertainty as a prior, UCMNet achieves state-of-the-art performance on multiple benchmarks with 30\% fewer parameters than previous models. Project page: \href{https://kdhrick2222.github.io/projects/UCMNet/}{https://kdhrick2222.github.io/projects/UCMNet}.

cs.CV↗

arXiv2026

ViKey: Enhancing Temporal Understanding in Videos via Visual Prompting

Recent advancements in Video Large Language Models (VideoLLMs) have enabled strong performance across diverse multimodal video tasks. To reduce the high computational cost of processing dense video frames, efficiency-oriented methods such as frame selection have been widely adopted. While effective at minimizing redundancy, these methods often cause notable performance drops on tasks requiring temporal reasoning. Unlike humans, who can infer event progression from sparse visual cues, VideoLLMs frequently misinterpret temporal relations when intermediate frames are omitted. To address this limitation, we explore visual prompting (VP) as a lightweight yet effective way to enhance temporal understanding in VideoLLMs. Our analysis reveals that simply annotating each frame with explicit ordinal information helps the model perceive temporal continuity. This visual cue also supports frame-level referencing and mitigates positional ambiguity within a sparsely sampled sequence. Building on these insights, we introduce ViKey, a training-free framework that combines VP with a lightweight Keyword-Frame Mapping (KFM) module. KFM leverages frame indices as dictionary-like keys to link textual cues to the most relevant frames, providing explicit temporal anchors during inference. Despite its simplicity, our approach substantially improves temporal reasoning and, on some datasets, preserves dense-frame baseline performance with as few as 20% of frames.

cs.CV↗

arXiv2026

DMD-augmented Unpaired Neural Schr\"odinger Bridge for Ultra-Low Field MRI Enhancement

Ultra Low Field (64 mT) brain MRI improves accessibility but suffers from reduced image quality compared to 3 T. As paired 64 mT - 3 T scans are scarce, we propose an unpaired 64 mT $\rightarrow$ 3 T translation framework that enhances realism while preserving anatomy. Our method builds upon the Unpaired Neural Schr\"odinge Bridge (UNSB) with multi-step refinement. To strengthen target distribution alignment, we augment the adversarial objective with DMD2-style diffusion-guided distribution matching using a frozen 3T diffusion teacher. To explicitly constrain global structure beyond patch-level correspondence, we combine PatchNCE with an Anatomical Structure Preservation (ASP) regularizer that enforces soft foreground background consistency and boundary aware constraints. Evaluated on two disjoint cohorts, the proposed framework achieves an improved realism structure trade-off, enhancing distribution level realism on unpaired benchmarks while increasing structural fidelity on the paired cohort compared to unpaired baselines.

cs.CV↗

arXiv2026

LittleBit-2: Maximizing the Spectral Energy Gain in Sub-1-Bit LLMs via Latent Geometry Alignment

We identify the Spectral Energy Gain in extreme model compression, where low-rank binary approximations outperform tiny-rank floating-point baselines for heavy-tailed spectra. However, prior attempts fail to realize this potential, trailing state-of-the-art 1-bit methods. We attribute this degradation to Latent Geometry Misalignment: standard singular vectors exhibit high coherence (spiky distribution), the worst-case geometry for binary quantization. To realize this gain, we propose LittleBit-2, a framework employing Internal Latent Rotation and Joint Iterative Quantization (Joint-ITQ). This approach acts as a geometric preconditioner, aligning coherent latent distributions with the binary hypercube with zero inference overhead. Empirically, LittleBit-2 establishes a new state-of-the-art in the sub-1-bit regime (1$\sim$0.1 bpp) on Llama-2 and Llama-3, matching the fidelity of leading 1-bit baselines.

cs.LG↗

arXiv2026

RaBiT: Residual-Aware Binarization Training for Accurate and Efficient LLMs

Efficient deployment of large language models (LLMs) requires extreme quantization, forcing a critical trade-off between low-bit efficiency and performance. Residual binarization enables hardware-friendly, matmul-free inference by stacking binary ($\pm$1) layers, but is plagued by pathological feature co-adaptation. We identify a key failure mode, which we term inter-path adaptation: during quantization-aware training (QAT), parallel residual binary paths learn redundant features, degrading the error-compensation structure and limiting the expressive capacity of the model. While prior work relies on heuristic workarounds (e.g., path freezing) that constrain the solution space, we propose RaBiT, a novel quantization framework that resolves co-adaptation by algorithmically enforcing a residual hierarchy. Its core mechanism sequentially derives each binary path from a single shared full-precision weight, which ensures that every path corrects the error of the preceding one. This process is stabilized by a robust initialization that prioritizes functional preservation over mere weight approximation. RaBiT redefines the 2-bit accuracy-efficiency frontier: it achieves state-of-the-art performance, rivals even hardware-intensive Vector Quantization (VQ) methods, and delivers a $4.49\times$ inference speed-up over full-precision models on an RTX 4090. Code is available at https://github.com/SamsungLabs/RaBiT.

cs.AI↗

arXiv2025

SlumpGuard: An AI-Powered Real-Time System for Automated Concrete Slump Prediction via Video Analysis

Concrete workability is essential for construction quality, with the slump test being the most widely used on-site method for its assessment. However, traditional slump testing is manual, time-consuming, and highly operator-dependent, making it unsuitable for continuous or real-time monitoring during placement. To address these limitations, we present SlumpGuard, an AI-powered vision system that analyzes the natural discharge flow from a mixer-truck chute using a single fixed camera. The system performs automatic chute detection, pouring-event identification, and video-based slump classification, enabling quality monitoring without sensors, hardware installation, or manual intervention. We introduce the system design, construct a site-replicated dataset of over 6,000 video clips, and report extensive evaluations demonstrating reliable chute localization, accurate pouring detection, and robust slump prediction under diverse field conditions. An expert study further reveals significant disagreement in human visual estimates, highlighting the need for automated assessment.

cs.CV↗

arXiv2025

A 2-Stage Model for Vehicle Class and Orientation Detection with Photo-Realistic Image Generation

We aim to detect the class and orientation of a vehicle by training a model with synthetic data. However, the distribution of the classes in the training data is imbalanced, and the model trained on the synthetic image is difficult to predict in real-world images. We propose a two-stage detection model with photo-realistic image generation to tackle this issue. Our model mainly takes four steps to detect the class and orientation of the vehicle. (1) It builds a table containing the image, class, and location information of objects in the image, (2) transforms the synthetic images into real-world images style, and merges them into the meta table. (3) Classify vehicle class and orientation using images from the meta-table. (4) Finally, the vehicle class and orientation are detected by combining the pre-extracted location information and the predicted classes. We achieved 4th place in IEEE BigData Challenge 2022 Vehicle class and Orientation Detection (VOD) with our approach.

cs.CV↗

arXiv2025

Speaking Beyond Language: A Large-Scale Multimodal Dataset for Learning Nonverbal Cues from Video-Grounded Dialogues

Nonverbal communication is integral to human interaction, with gestures, facial expressions, and body language conveying critical aspects of intent and emotion. However, existing large language models (LLMs) fail to effectively incorporate these nonverbal elements, limiting their capacity to create fully immersive conversational experiences. We introduce MARS, a multimodal language model designed to understand and generate nonverbal cues alongside text, bridging this gap in conversational AI. Our key innovation is VENUS, a large-scale dataset comprising annotated videos with time-aligned text, facial expressions, and body language. Leveraging VENUS, we train MARS with a next-token prediction objective, combining text with vector-quantized nonverbal representations to achieve multimodal understanding and generation within a unified framework. Based on various analyses of the VENUS datasets, we validate its substantial scale and high effectiveness. Our quantitative and qualitative results demonstrate that MARS successfully generates text and nonverbal languages, corresponding to conversational input.

cs.AI↗

arXiv2025

LittleBit: Ultra Low-Bit Quantization via Latent Factorization

The deployment of large language models (LLMs) is frequently hindered by prohibitive memory and computational requirements. While quantization mitigates these bottlenecks, maintaining model fidelity in the sub-1-bit regime remains a persistent challenge. In this paper, we introduce LittleBit, a novel framework for extreme LLM compression. We target quantization rates as low as $0.1$ bits per weight (BPW), achieving a memory reduction of approximately $31\times$, which effectively compresses Llama2-13B to under $0.9$ GB. We represent weights via low-rank latent matrix factorization and subsequently binarize the resulting factors. To counteract the information loss inherent to such drastic precision reduction, we integrate a multi-scale compensation mechanism that learns importance parameters across row, column, and latent dimensions. Two primary contributions enable effective training: Dual Sign-Value-Independent Decomposition (Dual-SVID) for quantization-aware training (QAT) initialization, and Residual Compensation to minimize approximation errors. Extensive experiments confirm the superiority of LittleBit in the sub-1-bit domain; for instance, our method at $0.1$ BPW surpasses the performance of leading techniques operating at $0.7$ BPW on Llama2-7B. We establish a new size-performance trade-off -- unlocking a potential $11.6\times$ inference speedup relative to FP16 -- and render powerful LLMs practical for resource-constrained environments. Our code is available at https://github.com/SamsungLabs/LittleBit.

cs.LG↗

arXiv2025

MAVL: A Multilingual Audio-Video Lyrics Dataset for Animated Song Translation

Lyrics translation requires both accurate semantic transfer and preservation of musical rhythm, syllabic structure, and poetic style. In animated musicals, the challenge intensifies due to alignment with visual and auditory cues. We introduce Multilingual Audio-Video Lyrics Benchmark for Animated Song Translation (MAVL), the first multilingual, multimodal benchmark for singable lyrics translation. By integrating text, audio, and video, MAVL enables richer and more expressive translations than text-only approaches. Building on this, we propose Syllable-Constrained Audio-Video LLM with Chain-of-Thought SylAVL-CoT, which leverages audio-video cues and enforces syllabic constraints to produce natural-sounding lyrics. Experimental results demonstrate that SylAVL-CoT significantly outperforms text-based models in singability and contextual accuracy, emphasizing the value of multimodal, multilingual approaches for lyrics translation.

cs.CL↗

arXiv2025

Mantra: Rewriting Quantum Programs to Minimize Trap-Movements for Zoned Rydberg Atom Arrays

A zoned neutral atom architecture achieves exceptional fidelity by segregating the execution spaces of 1- and 2-qubit gates, being a promising candidate for high-accuracy quantum systems. Unfortunately, naively applying programs designed for static qubit topologies to zoned architectures may result in most execution time being consumed by inter-zone travels of atoms. To address this, we introduce Mantra (Minimizing trAp movemeNts for aTom aRray Architectures), which rewrites quantum programs to reduce the interleaving of single- and two-qubit gates. Mantra incorporates three strategies: (i) a fountain-shaped controlled-Z (CZ) chain, (ii) ZZ-interaction protocol without a 1-qubit gate, and (iii) preemptive gate scheduling. Mantra reduces inter-zone movements by 68%, physical gate counts by 35%, and improves circuit fidelities by 17% compared to the standard executions.

quant-ph↗

arXiv2025

Balancing Thermal Relaxation Deviations of Near-Future Quantum Computing Results via Bit-Inverted Programs

One of the predominant causes of program distortion in the real quantum computing system may be attributed to the probability deviation caused by thermal relaxation. We introduce Barber (Balancing reAdout Results using Bit-invErted ciRcuits), a method designed to counteract the asymmetric thermal relaxation deviation and improve the reliability of near-term quantum programs. Barber collaborates with a bit-inverted quantum circuit, where the excited quantum state of qubits is assigned to the $\lvert 0 \rangle$ and the unexcited state to the $\lvert 1 \rangle$. In doing so, bit-inverted quantum circuits can experience thermal relaxation in the opposite direction compared to standard quantum circuits. Barber can effectively suppress the thermal relaxation deviation in program's readout results by selectively merging distributions from the standard and bit-inverted circuits.

quant-ph↗

arXiv2024

Distribution-Adaptive Dynamic Shot Optimization for Variational Quantum Algorithms

Variational quantum algorithms (VQAs) have attracted remarkable interest over the past few years because of their potential computational advantages on near-term quantum devices. They leverage a hybrid approach that integrates classical and quantum computing resources to solve high-dimensional problems that are challenging for classical approaches alone. In the training process of variational circuits, constructing an accurate probability distribution for each epoch is not always necessary, creating opportunities to reduce computational costs through shot reduction. However, existing shot-allocation methods that capitalize on this potential often lack adaptive feedback or are tied to specific classical optimizers, which limits their applicability to common VQAs and broader optimization techniques. Our observations indicate that the information entropy of a quantum circuit's output distribution exhibits an approximately exponential relationship with the number of shots needed to achieve a target Hellinger distance. In this work, we propose a distribution-adaptive dynamic shot (DDS) framework that efficiently adjusts the number of shots per iteration in VQAs using the entropy distribution from the prior training epoch. Our results demonstrate that the DDS framework sustains inference accuracy while achieving a ~50% reduction in average shot count compared to fixed-shot training, and ~60% higher accuracy than recently proposed tiered shot allocation methods. Furthermore, in noisy simulations that reflect the error rates of actual IBM quantum systems, DDS achieves approximately a ~30% reduction in the total number of shots compared to the fixed-shot method with minimal degradation in accuracy, and offers about ~70% higher computational accuracy than tiered shot allocation methods.

quant-ph↗