Search arXiv⌕ Search

arXiv subjects

Ke Chen

Publications and source records attributed to Ke Chen.

At least 19 recordsLinked to original sources
arXiv2026

DICE: Decentralized Incentive-driven Federated Learning for Efficient & Scalable Customization of Pretrained Models in Web3

Customizing Pretrained Models (PMs) for downstream tasks demands task-specific data, which burgeoning Web3 applications provide abundantly. Blockchain-powered federated learning (BCFL) offers a secure and privacy-preserving paradigm to customize PMs among self-interested and untrusted Web3 participants. Yet, existing BCFL algorithms face severe efficiency and scalability challenges when applied to large PMs, due to their substantial computational and memory requirements and the hardware constraints of many participants. This paper introduces DICE, a novel BCFL algorithm that addresses the efficiency and scalability challenges of customizing large PMs in Web3 via two sets of synergistic techniques: (i) a dynamic progressive learning scheme and a model slice verification scheme to enable a wider range of honest participants, including those unable to deploy the full model, to engage in customizing PMs, while preserving model utility; (ii) a tree-structured committee consensus protocol and an efficiency-driven incentive scheme to address the dominant bottleneck that hinders existing BCFL algorithms from scaling to a larger number of data holders for customizing data-hungry large PMs, while maintaining security. Theoretical analysis and extensive experimental results on customizing vision and language PMs demonstrate DICE's efficiency, effectiveness, and security.

cs.CR↗
arXiv2026

CrossEdit: Cross-Modal Training Enables Rich Audio-Visual Editing

Multimedia editing requires generative models to interpret complex, compositional instructions and modify only the desired elements across modalities, a capability largely beyond existing systems. Current approaches are typically limited to simple, single-attribute edits within one modality, since diverse, high-quality editing pairs are hard to obtain, especially for cross-modal tasks where aligned audiovisual (AV) data is scarce. We present CrossEdit, a unified omni-modal editing model for images, audio, and video that performs AV movie scene edits zero-shot through cross-modal transfer. Our key observation is that complex instruction-following, once learned in any modality, generalizes to others. We exploit the additive nature of acoustic signals to procedurally generate large-scale audio editing pairs with compositional instructions, and show that fine-tuning on this synthetic data and self-supervised AV masked reconstruction, alongside a targeted set of curated cross-modal tasks, induces instruction-following that transfers zero-shot to unseen modality and instruction combinations. To evaluate this capability, we release CrossEditBench, a human-annotated benchmark of AV edits on movie scenes, and propose AV-FES, a metric that jointly scores instruction following and consistency. We show that our proposed techniques improve performance on zero-shot AV movie scene editing, while maintaining or improving performance on lip-synced speech editing and standard image, video, and audio editing benchmarks. Demos are available at https://wanchichen.github.io/crossedit/.

cs.SD↗
arXiv2026

TAC2: Timestamped Audio Captioning as composable functions

Real-world audio applications require precise, queryable representations grounded in physical time. Yet modern Large Audio-Language Models (LALMs) increasingly favor conversational generation, producing free-form descriptions that conflate acoustic perception with higher-level reasoning and lack precise temporal grounding. We introduce TAC2, a foundation model that casts timestamped audio captioning as composable functions to construct granular, speaker-attributed acoustic timelines directly from in-the-wild audio. To avoid prohibitively expensive manual annotation, we programmatically fuse complementary predictions from specialized weak teachers through time-registered IoU matching. We model the resulting multi-granularity schema of timestamped events, speaker turns, and fine-grained speech attributes using an Instruction-Input-Output (IIO) completion format, in which each task is a function of the audio, an instruction, and an optional input. A single call decodes the full description autonomously, while composing calls lets earlier outputs or external priors, such as a reference transcript, condition later ones. We further extend TAC2 with expressive speech attributes to obtain TAC2-S, using parameter-efficient specialists to elicit latent paralinguistic capabilities from the pre-trained backbone and distill them into a unified model. We validate the resulting representation in two complementary settings. Paired with frozen text-only LLMs, it yields strong spoken audio understanding across diverse reasoning backends, demonstrating the utility of decoupling fine-grained acoustic perception from downstream reasoning. Used as paired supervision for text-to-audio generation, it supports semantically and temporally grounded reconstruction, together with explicit zero-shot control over event timing, speaker attributes, and word-level emphasis.

cs.SD↗
arXiv2026

RAG-CSI: Global Local Retrieval via Joint State and Dynamics Matching for CSI Prediction

Channel state information (CSI) prediction forecasts future channel states from historical observations, supporting transceiver design and resource allocation with reduced measurement overhead. However, existing deep learning predictors mainly encode channel knowledge in model parameters, which can limit their ability to represent diverse patterns of channel dynamics across scenes. In this paper, we propose RAG-CSI, a retrieval-enhanced Transformer framework that augments prediction with relevant historical channel knowledge. Specifically, we construct a global--local CSI knowledge store containing representative observations across scenes and historical observations from the current serving scene. We further develop joint state--dynamics retrieval (JSDR), which jointly characterizes channel states, phase dynamics, and amplitude dynamics to retrieve relevant CSI sequences. A Transformer predictor then uses cross-attention to fuse the retrieved knowledge with the current CSI history and predict future CSI with limited additional model parameters. Experiments on a multi-scene CSI dataset show overall improvements exceeding 10\% across all five evaluation metrics over the best baseline for each metric, including a 15.42\% reduction in normalized mean squared error.

eess.SP↗
arXiv2026

Superpixel-Based Image Segmentation Using Squared 2-Wasserstein Distances

We present an efficient method for image segmentation in the presence of strong intensity inhomogeneities. The approach combines two levels of clustering: pixels are first grouped into superpixels using Power-SLIC, which admits a discrete optimal transport (OT) interpretation, and these superpixels are subsequently merged using squared 2-Wasserstein distances between their empirical distributions. We formulate the merging process as an adjacency-constrained hierarchical optimization problem with a regularized objective that accounts for the heterogeneities of previously formed regions. We establish NP-completeness of the associated global optimization problem, even for two final regions and two-bin histograms. For the proposed greedy algorithm, we prove exact recovery under suitable histogram separation conditions and derive an additive approximation bound for the regularized energy. Numerical experiments demonstrate competitive segmentation accuracy on images exhibiting substantial intensity inhomogeneity, while retaining computational efficiency.

cs.CV↗
arXiv2026

KunSet: A Corpus and Benchmark for Lyrics Alignment and Rhyme Retrieval in Kunqu Opera

Kunqu is one of the oldest extant forms of Chinese opera, yet its digitization remains scarce and difficult to annotate, with two challenges standing out: first, close analysis of Kunqu performance demands syllable-level lyrics alignment; second, analyzing Kunqu pronunciation requires its phonological attributes (e.g., rhyme) to be linked with each syllable. To address these challenges, we present KunSet, the first Kunqu performance corpus comprising 482,164 fine-grained annotations of syllable-level lyrics and phonological attributes in total media duration of 155.2 h with 106.3 h annotated. To ensure annotation quality, KunSet was preannotated by an iterative ML-assisted pipeline and corrected by over 200 trained annotators via a purpose-built online annotation system under a two-round review process. To assess KunSet as a music information retrieval (MIR) benchmark, we evaluate two tasks: lyrics alignment, by finetuning a forced-alignment model with an acoustic anchor decoder; and same-rhyme retrieval, using frozen Chinese HuBERT features. Models built with KunSet outperform their baselines. We release KunSet and model checkpoints to support MIR research on traditional-music corpora.

cs.SD↗
arXiv2026

Data Completion for Electrical Impedance Tomography by Conditional Diffusion Models

Data scarcity is a fundamental barrier in Electrical Impedance Tomography (EIT), as undersampled Dirichlet-to-Neumann (DtN) measurements can substantially degrade conductivity reconstructions. We address this bottleneck by completing partially observed DtN measurements using a diffusion based generative model. Specifically, we train a conditional diffusion model to learn the distribution of DtN data and to infer full measurement vectors given partial observations. Our approach supports flexible source receiver configurations and can be used as a plug in preprocessing step with off the shelf EIT solvers. Under mild assumptions on the polygon conductivity class, we derive nonasymptotic end to end bounds on the distributional discrepancy between the completed and ground truth DtN measurements. In numerical experiments, we couple the proposed diffusion completion procedure with a deep learning based inverse solver and compare its performance against the same solver with full measurement data. The results show that diffusion completion enables reconstructions comparable to the full data baseline while using only 1% of the measurements. In contrast, standard baselines such as matrix completion require 30% of the measurements to achieve similar reconstruction quality.

math.NA↗
arXiv2026

Hexagonal Basis Projectors for Model-Based Iterative CT Reconstruction

Image representation affects both the forward model and quantitative feature recovery in model-based iterative CT reconstruction. For isotropically bandlimited signals, the hexagonal lattice is the most sampling-efficient lattice in two dimensions, with greater directional isotropy and a larger alias-free bandwidth than the Cartesian grid at equal density. We develop constant and linear hexagonal projectors for two-dimensional fan-beam CT using box spline calculus. Both bases form a partition of unity on common coefficient centers; the linear tent also reproduces affine functions. We derive analytic ray footprints and effective detector-blur expressions and implement efficient GPU forward and backprojection. On simulated FORBILD data, the linear model improves on Cartesian and hexagonal pixels, with its best noisy-data accuracy at the coarser resolution. Real-world 2DeteCT comparisons use high-dose, high-resolution Cartesian reconstructions as pseudo ground truth and show competitive image and feature agreement. These results establish a practical hexagonal representation for quantitative CT.

cs.CV↗
arXiv2026

Scaling-Critical Theory for the Boltzmann and Landau Equations

This paper develops a scaling-critical theory for the spatially inhomogeneous non-cutoff Boltzmann and Landau equations with very soft potentials. For sufficiently small initial perturbations in a localized, weighted, anisotropic Riesz-potential norm, we prove global well-posedness near a Maxwellian in the whole space. This critical phase-space norm captures the Boltzmann--Landau scaling, the velocity-dependent anisotropy, and the hypoelliptic transport structure. The proof combines frozen-operator estimates, a critical fixed-point argument, and weighted hypocoercive energy estimates. We also develop a short-time pointwise Green-function theory for variable-coefficient kinetic equations with a nonnegative Hölder background. We first construct the small-jump Green function by freezing coefficients along kinetic characteristics and then recover the full kernel through a convergent parametrix expansion. The resulting bounds capture the fractional Kolmogorov geometry near characteristics and rapid decay away from them. These estimates form the analytic core of the scaling-critical theory and provide a robust pointwise framework for further regularity analysis in the soft-potential regime.

math.AP↗
arXiv2026

Optimization of gridding algorithms for FFT by vector optimization

The Fast Fourier Transform (FFT) is widely used in applications such as MRI, CT, and interferometry; however, because of its dependence on uniformly sampled data, it requires the use of gridding techniques for practical implementation. The performance of these algorithms strongly depends on the choice of the gridding kernel, with the first prolate spheroidal wave function (PSWF) regarded as optimal. This work redefines kernel optimality through the lens of vector optimization (VO), introducing a rigorous framework that characterizes optimal kernels as Pareto-efficient solutions of an error shape operator. We establish the continuity of such operator, study the existence of solutions, and propose a novel methodology to construct kernels tailored to a desired target error function. The approach is implemented numerically via interior-point optimization. Comparative experiments demonstrate that the proposed kernels outperform both the PSWF and the state-of-the-art methods (MIRT-NUFFT) in specific regions of interest, achieving orders-of-magnitude improvements in mean absolute errors. These results confirm the potential of VO-based kernel design to provide customized accuracy profiles aligned with application-specific requirements. Future research will extend this framework to multidimensional cases and relative error minimization, with potential integration of machine learning for adaptive target error selection.

math.NA↗
arXiv2026

PromptSep-Speaker: Generative Target Speaker Extraction via Multimodal Prompting and Semantic Alignment

Target speaker extraction (TSE) isolates a desired speaker from a multi-speaker mixture using cues that identify the target. Prior work has explored acoustic and textual cues, such as enrollment audio and different types of speaker descriptors, but typically relies on a single cue family. Yet no single cue is both broadly available and reliably discriminative. Moreover, most TSE systems adopt mask-prediction models, which are prone to audible artifacts and generalize poorly across acoustic conditions. To address these limitations, we propose PromptSep-Speaker, a generative TSE model based on latent diffusion that unifies four interchangeable cues: speaker temporal order, transcript, enrollment audio, and speaker caption. The generative formulation delivers higher separation quality and multi-modal prompting improves applicability across scenarios. To fully exploit each cue, we further introduce a frozen WavLM representation with an auxiliary alignment loss that supervises query-mixture cross-attention without extra labeling. Across four evaluation sets and a listening study, PromptSep-Speaker achieves state-of-the-art performance over TSE and universal source separation baselines, improving separation accuracy and quality across cue modalities.

cs.SD↗
arXiv2026

Uncertainty-Aware Probabilistic Constrained Clustering from Entangled Pairwise Supervision

Pairwise constrained clustering typically relies on hard must-link/cannot-link labels, whereas realistic pairwise supervision may be real-valued and entangle intrinsic ambiguity, expert judgment, and stochastic corruption. Existing deep constrained clustering (DCC) methods mainly target hard, expert-agnostic constraints, treating soft labels mostly numerically rather than semantically. We formalize this setting as uncertainty-aware probabilistic constrained clustering (UPCC), defining a canonical aleatoric relation through a heterogeneous observation process and analyzing its conditional identifiability. We introduce ProbPair, an angular pairwise objective for probabilistic relations, and build ECI-PP, an estimator--corrector--integrator framework that refines imperfect supervision via belief estimation, correction, and reliability-aware integration. Across challenging probabilistic supervision settings, experiments on diverse benchmarks show that ECI-PP outperforms state-of-the-art DCC methods and remains robust with a shared default configuration.

cs.LG↗
arXiv2026

EGCForest: Structured Agent Memory Organization for Enterprise Group Chats

Enterprise group chats record project progress, decisions, task ownership, and other organizational information valuable to enterprise agents. Reliable retrieval is challenging because evidence spans interleaved threads, channels, and time, with meaning and authority conditioned on the speaker. Neither RAG's independent chunk retrieval nor memory systems' summary-based consolidation can reliably recover such distributed evidence. We present EGCForest, a novel multi-view hierarchical memory framework for enterprise group chats. EGCForest organizes an immutable message log into three complementary forests: a Channel forest for temporal workflows, an Actor forest for speaker-conditioned context, and an Artifact forest for deliverable lifecycles. Cross-view connections are enabled by a Placement Map and sparse structural bridges. Built on this memory backbone, EGCForest retrieves and assembles evidence by jointly leveraging message-level relevance, hierarchical context from forest summaries, and cross-view links, recovering distributed evidence. Experiments on three enterprise-chat benchmarks show that EGCForest consistently outperforms RAG and memory-system baselines, improving answer accuracy by up to 20.85 points over MemOS while supporting efficient online retrieval.

cs.CL↗
arXiv2026

Prompt Stability Matters: Evaluating and Optimizing Auto-Generated Prompt in General-Purpose Systems

Automatic prompt generation plays a crucial role in enabling general-purpose multi-agent systems to perform diverse tasks autonomously. Existing methods typically evaluate prompts based on their immediate task performance, overlooking the intrinsic qualities that determine their reliability. This outcome-centric view not only limits interpretability but also fails to account for the inherent stochasticity of large language models (LLMs). In this work, we bring attention to prompt stability-the consistency of model responses across repeated executions-as a key factor for building robust and effective prompt generation systems. To quantify this, we propose semantic stability as a criterion for assessing the response consistency of prompts. Based on the proposed metric, we developed the first stability-aware general-purpose prompt generation system that leverages stability feedback to iteratively enhance both prompt quality and system-level performance. Furthermore, we establish a logical chain between prompt stability and task success by analyzing the structural dependencies within our system, proving stability as a necessary condition for effective system-level execution. Empirical results across general and domain-specific tasks demonstrate that our stability-aware framework improves both accuracy and output consistency. By shifting the focus from one-off results to persistent reliability, our work offers a new perspective on prompt design and contributes practical tools for building more trustworthy general-purpose systems. The UI and source code for our system are publicly available at: https://xucheng63.github.io/MyDataPilot/.

cs.AI↗
arXiv2026

LATTEArena: An Evaluation Framework for LLM-powered Tabular Feature Engineering (Extended Version)

Feature engineering remains a cornerstone of tabular data analysis, and Large Language Models (LLMs) have emerged as a promising paradigm for its automation, giving rise to LLM-powered Automated Tabular Feature Engineering (LATTE). However, the field lacks standardized, cost-aware evaluation platforms, and the combinatorial explosion of design choices obscures true algorithmic progress. To bridge these gaps, we systematically deconstruct 15 representative LATTE methods into a unified 6-dimensional taxonomy. Based on this abstraction, we introduce LATTEArena, a standardized, modular, and extensible benchmarking framework that decouples monolithic pipelines into reusable execution blocks. By distilling the massive combinatorial space, we evaluate 24 core LATTE configurations across 7 research questions. Our head-to-head benchmarking goes beyond predictive accuracy to quantify token efficiency and execution robustness, yielding 17 empirical findings on cost-effectiveness trade-offs. Furthermore, we provide 3 concrete recommendations for optimal real-world deployment. By enabling controlled component-level comparisons, LATTEArena shifts the paradigm from ad-hoc prompt engineering to systematic context management. All code, datasets, and over 4,000 execution logs are publicly available to foster a dynamic, community-driven benchmark. Our framework, leaderboard, and all artifacts are hosted on the LATTEArena project website at https://goodenhak.github.io/LATTEArena.

cs.AI↗
arXiv2026

TASTE: Taxonomy-Aware Symbolic Piano Generation from Textual Description

Text-conditioned symbolic music generation has attracted growing interest with the development of generative models. Yet two questions remain underexplored: what aspects of music the conditioning text should describe, and how the corresponding data should be collected and trained to improve controllability and text-music alignment. We address these challenges with TASTE, a taxonomy-aware framework for symbolic music generation with a main focus on piano pieces. First, we introduce a taxonomy of musical elements that specifies how a piece can be captured through textual description in a musically grounded way. The taxonomy comprises a set of metadata tags (genre, emotion, section plan, etc.) together with a free-form description of music imagery. This structure decomposes textual input to match the generation goal more coherently and precisely. Second, we build a data pipeline for collecting, annotating, and retrieving public symbolic piano resources, yielding a corpus of 432K MIDI files labeled at multiple levels of granularity. Finally, we design a staged training procedure that leverages this multi-granularity data across two model families: an autoregressive model and a non-autoregressive diffusion model. Both objective and subjective evaluations show that our models achieve superior musical-element control, text-description alignment, and overall music quality. We will release the full pipeline, all annotated datasets, and the model checkpoints.

cs.SD↗
arXiv2026

MedMemoryBench: Benchmarking Agent Memory in Personalized Healthcare

The large-scale deployment of personalized healthcare agents demands memory mechanisms that are exceptionally precise, safe, and capable of long-term clinical tracking. However, existing benchmarks primarily focus on daily open-domain conversations, failing to capture the high-stakes complexity of real-world medical applications. Motivated by the stringent production requirements of an industry-leading health management agent serving tens of millions of active users, we introduce MedMemoryBench. We develop a human-agent collaborative pipeline to synthesize highly realistic, long-horizon medical trajectories based on clinically grounded, synthetic patient archetypes. This process yields a massive, expertly validated dataset comprising approximately 2,000 sessions and 16,000 interaction turns. Crucially, MedMemoryBench departs from traditional static evaluations by pioneering an "evaluate-while-constructing" streaming assessment protocol, which precisely mirrors dynamic memory accumulation in production environments. Furthermore, we formalize and systematically investigate the critical phenomenon of memory saturation, where sustained information influx actively degrades retrieval and reasoning robustness. Comprehensive benchmarking reveals severe bottlenecks in mainstream architectures, particularly concerning complex medical reasoning and noise resilience. By exposing these fundamental flaws, MedMemoryBench establishes a vital foundation for developing robust, production-ready medical agents.

cs.AI↗
arXiv2026

CORAL: Cross-modal Vector Retrieval via Incremental Graph Construction at Scale

Cross-modal vector retrieval is widely used in multimodal systems, such as search engines and vector databases. It typically operates in out-of-distribution (OOD) settings, where query vectors follow a distribution that differs from that of the vectors stored in the database. In such cases, conventional indexes suffer significant performance degradation, and even methods specially designed for OOD remain limited by inefficient use of query modal characteristics, restricted GPU parallelism, and inadequate support for dynamic updates. We present CORAL, a novel GPU-accelerated graph-based vector index for scalable cross-modal retrieval, featuring hierarchical memory management that spans GPU, CPU, and disk. Specifically, CORAL incrementally incorporates the characteristics of query modality and terminates index construction timely. Crucially, it introduces coverage-aware adaptive pruning to address the imbalanced coverage of the query vector's neighbors. Moreover, CORAL presents a fully neighborhood-aware projection approach to efficiently utilize GPUs for highly parallel index construction, and a targeted connectivity enhancement method to refine the index structure. Besides, CORAL also supports modal-semantics-based vector insertion and topology-repairing deletion that restore node connectivity. Experimental results demonstrate that CORAL outperforms existing methods with up to 1.6 times the throughput at matched recall while reducing construction time by up to 56%. Furthermore, it exhibits remarkable resilience under dynamic updates and remains effective at the billion scale.

cs.DB↗