跳到正文
研踪PaperTrail

论文库

检索计算机领域论文,按研究方向、作者或关键词筛选。

部分采集异常
正在读取账号与研究记录…

全站更新概况

+197新增论文
+151其中 AI 相关
+16已有论文修订
+91新增 AI 导读

arXiv 最近完成扫描:2026-11-21 12:19 · 全站最近入库:2026-11-21 12:19各信源进度 →

统计口径与数据新鲜度

统计 2026-11-21 05:33 — 2026-11-22 05:33(北京时间) · 点击数字查看对应简报;AI 相关包含在新增论文内。

最近完成检查:2026-11-21 12:19最近发现新增版本:2026-11-21 12:19库内论文最新更新时间:2026-11-20 02:59

每分钟刷新已保存的数据。空检查不代表采集异常;新收录可能包含历史论文。AI 导读按新增版本计数。

29,158 已收录论文24,181 AI 相关8,992 有修订112 HF 社区发现

浏览论文库

时间范围跟随排序:版本更新、本站收录或首次发表。最新收录可能包含历史论文。部分信源异常,查看状态 →

找到 21,490 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-21 12:17(北京时间)
版本更新 · v2arXiv:2606.08678

Speaker-Invariant Representation Learning for Spoofing Detection via Gradient Reversal and A Variational Information Bottleneck

Anh-Tuan Dao, Driss Matrouf, Mickael Rouvier, Nicholas Evans

原文摘要

Sophisticated generative speech technology can undermined the reliability of voice biometrics. While spoofing detection systems excel when assessed under in-domain conditions, generalisation to out-of-domain settings is often poor. In this paper, we show that such issues could be caused by speaker bias, where models learn individual voice traits rather than markers of manipulation or generation. We propose a teacher-student framework for speaker-invariant spoofing detection that disentangles identity without requiring speaker labels. We leverage a pre-trained speaker recognition teacher to guide a student model via a gradient reversal layer. To control the balance between suppressing cues related to voice identity with the preservation of those related to spoofing detection, we integrate a Variational Information Bottleneck. Evaluations across nine datasets show our model achieves a 25.7% relative reduction to the EER compared to the MHFA baseline.

cs.SD · cs.LG

版本更新 · v2arXiv:2610.11161

VGGT-VLA-World: Intent-Conditioned 3D World Evolution for Autonomous Driving

Zhaoyang Liu, Zhigui Lin, Kun Jiang, Tianhui Li, Changye Liu, Ziying Song, Diange Yang

原文摘要

VGGT provides a strong foundation for geometry-centric world models by recovering unified 3D scene geometry from visual observations. Although recent extensions enable temporal 3D prediction, their future evolution remains weakly conditioned on driving intentions and actions, limiting their ability to model alternative action-dependent futures. We propose \textbf{VGGTWorld-VLA}, an intention-conditioned extension of VGGT-World for controllable 3D world evolution in autonomous driving. First, we introduce an action--semantic conditioning mechanism that injects complementary driving semantics and ego-motion representations into the future-token stream, enabling different future geometry predictions for the same observed scene under alternative ego actions. Second, we develop a geometry--language--action bridge that adapts historical geometry, VLA semantic features, and maneuver and trajectory representations for joint conditioning of future geometry prediction. We evaluate future geometry prediction on NAVSIM, with ablation studies examining the contributions of semantic and action conditioning. Compared with baseline, our method reduces future-depth RMSE by 13.7\% and improves $δ_1$ by 4.11 percentage points on NAVSIM motion subset. Controlled action-input interventions further show that the model uses the supplied action information. These results support semantic and action conditioning for controllable VGGT-based world prediction in autonomous driving.

cs.CV · cs.RO

新论文 · v1arXiv:2611.21165

A Matter of Stages: Energy-Latency Trade-offs in Multimodal LLM Serving

Yifan Sun, Jiantong Jiang, Zhexiang Zhang, Xu Bai, Feng Liu, Rajkumar Buyya, Adel N. Toosi

原文摘要

Online serving of multimodal large language models requires deciding whether to separate visual encoding (E), prefill (P), and decode (D) across GPUs and how to configure their frequencies. EPD disaggregation enables independent stage control, but it also changes communication and queueing behavior. Lowering one stage's frequency can increase queueing and prolong execution across the EPD pipeline. Consequently, reducing power at one stage does not necessarily reduce total GPU energy per request. In this paper, we systematically characterize fused and EPD-disaggregated serving across four multimodal models under a fixed three-GPU budget, followed by single-stage and uniform downclocking experiments across 18 workload combinations. We find that disaggregation reduces GPU energy per successful request by 13.3-37.0% at nearly identical output-token throughput, but increases mean TTFT. Downclocking exhibits distinct stage sensitivities, and aggressive frequency reductions can offset power savings through queueing and longer execution. Across the evaluated workloads, uniform downclocking to 80% reduces GPU energy by 11.7-18.6% with modest mean-latency changes. Guided by these findings, we develop a budgeted frequency exploration strategy to select energy-efficient configurations under latency constraints. Evaluation shows that four probes can identify the lowest-energy feasible configuration among 25 measured candidates across all 288 evaluated positive-margin SLO scenarios. These results provide practical guidance for energy-efficient multimodal serving.

cs.DC

版本更新 · v6arXiv:2605.18267

SRC-Flow: Compact Semantic Representations Enable Normalizing Flows for Image Generation

Longtao Jiang, Jianmin Bao, Zhendong Wang, Xin Tao, Zhihui Li, Xiaojun Chang

原文摘要

Normalizing flows (NFs) provide exact likelihoods and deterministic invertible sampling, but have historically lagged behind diffusion models for large-scale image generation. We identify a key obstacle: NFs are required to learn a single invertible transport over the full ambient space, making them highly sensitive to high-dimensional representations. This leads to a semantic-capacity mismatch in modern visual representation spaces, where semantic information is compact but encoded in overcomplete features. We propose SRC-Flow, which introduces a Semantic Representation Compressor (SRC) to compact high-dimensional RAE features into a low-dimensional semantic space before flow modeling and preserve reconstruction through the frozen RAE decoder. This compact space reduces the modeling burden of NFs and enables effective likelihood-based generation in semantic representation space. We further adopt constant noise regularization tailored to the fixed unconditional bijection learned by flows. On ImageNet 256x256 and 512x512, SRC-Flow achieves state-of-the-art generation quality among normalizing flow methods, with gFID scores of 1.65 and 2.07 under classifier-free guidance, while retaining exact likelihood computation in the compact semantic representation space and deterministic invertible sampling at the flow level. Codes and models will be released in the future.

cs.CV

新论文 · v1arXiv:2611.21163

Polygraph: Large-Scale Multi-Session Trajectory Optimization using Pose Graphs

Petar Lukovic, Tian Yi Lim, Marc Pollefeys, Pablo Speciale, Hermann Blum, Zuria Bauer

原文摘要

Robots, vehicles, and wearable devices often traverse the same environment many times. Each pass produces its own trajectory from onboard visual or visual-inertial odometry. These trajectories are locally accurate but drift over time and live in unrelated coordinate frames. Turning many such independent trajectories into one globally consistent set is the large-scale multi-session trajectory optimization problem. Existing solutions largely work offline. All sessions must be available up front, one session is fixed as the reference frame, and a single fused map is then optimized. This paper presents Polygraph, a multi-session trajectory optimization pipeline that processes sessions online, one keyframe and one spatial cluster at a time, and never builds a 3D map. Keyframes are grouped into local clusters. When a cluster closes, it is matched against every other session through learned image retrieval. Candidate pairs are verified with sparse feature matching, monocular metric depth, and robust pose estimation. A two-tier consensus gate admits a cross-session edge into the shared pose graph only once independent evidence corroborates it. A cheap incremental solve keeps the graph current, and a periodic robust batch solve prunes outliers. We evaluate Polygraph on stereo outdoor driving, HoloLens and iPad indoor traversals, legged indoor locomotion with a Boston Dynamics Spot, and large-scale outdoor vehicle and handheld routes. Against the closest prior work on the same GPU model, using its reported timings, Polygraph runs faster, commits fewer loop-closure edges, and matches or improves on accuracy.

cs.RO

版本更新 · v2arXiv:2611.15413

MORA: Correcting Time Series Forecasts with Coarse Future Signals

Hu Li, Kenan Li, Haonan Xu, longlong, Shuqin Gao, Yucheng Zhang, Feng Dai

原文摘要

Recent architectural advances have improved time-series forecasting. However, our analysis of forecasters with diverse architectures reveals a common pattern: a substantial share of residual energy is concentrated in a small number of low-frequency components. Simple linear probes further show that the forecasters' historical inputs still contain predictive information about errors in the forecast mean. Building on these findings, we propose Model-Output Residual Adaptation (MORA), a model-agnostic residual adapter that refines the outputs of frozen forecasters without requiring additional inputs. MORA estimates segment-wise means of the target sequence through structural extrapolation and learned estimation, rather than directly predicting backbone-specific residuals. It then converts discrepancies between these estimates and the corresponding statistics of the backbone forecast into stagewise corrections, scaling each update according to estimation reliability. Extensive experiments on twelve benchmarks with six forecasting backbones show average MSE reductions of 7.2% for long-term forecasting and 28.1% for short-term forecasting, with lower MSE in 96.2% of the 288 evaluated cases. The benefits also extend to three time-series foundation models evaluated at different context lengths. Together, these results demonstrate that predictable coarse-scale residuals can be effectively exploited to improve existing forecasts. Code repository: https://github.com/hugerli-1/MORA.

cs.LG

新论文 · v1arXiv:2611.21162

DRaGAN: Dynamic Graph Attention Network for Multi-Radar Point Cloud Encoding in Automotive Radar Perception

Christof Leitgeb, Thomas Puchleitner, Max Peter Ronecker, Daniel Watzenig

原文摘要

Radar represents a key modality for weather-robust perception. Unlike optical sensors such as camera and LiDAR, radar can penetrate fog, rain and snow due to its longer wavelength, and recent hardware advances have steadily improved its spatial resolution, thus making it increasingly relevant for autonomous systems. While processing raw radar data requires substantial memory computation, most existing work has focused on preprocessed point-cloud data similar to LiDAR. However, the sparse radar points and sensor-relative anisotropy make LiDAR-oriented encoders suboptimal. To address this problem, we present DRaGAN, a dynamic graph attention network for multi-radar point clouds. DRaGAN introduces a sensor-relative Fourier encoding of angle and range to capture fixed-aperture radar characteristics, gated short- and long-range hub nodes along each sensor's line of sight for efficient local-global message passing, and attention-guided dynamic rewiring that prunes weak edges and adds a bounded number of latent connections restricted to two-hop neighborhoods. This reduces the number of processed edges in deeper layers while preserving representation quality. On TruckScenes, DRaGAN improves test NDS by 10.10 points over the evaluated RadarGNN baseline and reduces the reported model FLOP count by 63.7%. This trade-off includes an increase in model latency from 40 to 53.16 ms and in trainable parameters from 1.2M to 1.7M on the evaluated implementations. Code and pretrained models are available at https://github.com/chr-is-tof/DRaGAN.

cs.CV

新论文 · v1arXiv:2611.21160

PlainPairs: A Corpus Pairing the Prescriber and Patient Texts of U.S. Prescription Medicine Labels

Obada Kraishan

原文摘要

A prescription medicine in the United States comes with labeling written for prescribers, and many medicines also carry a text written for patients: a Medication Guide or a patient package insert. The two texts describe the same medicine for two readers, yet public resources rarely hold them side by side. I present PlainPairs, an open corpus built from the openFDA drug label files. Version 0.1.0 holds 35,177 prescription labels that carry a patient text, 21,416 distinct patient texts, 250,829 patient-text segments matched to prescriber sections through their regulatory headings, reading-level measures on both sides, and rule-based counts of plain-language wording. A second source, the National Cancer Institute's cancer information summaries, adds 756 summaries in health-professional and patient versions, in English and Spanish. Measured on the structure of the label, the median Medication Guide reads at grade 8.7 by Flesch-Kincaid and at grade 11.1 by SMOG. The same guides measured on the flattened text that the bulk files serve read about two grades higher, a difference that comes from lost list and heading boundaries. Patient texts address the reader in 41 of every 100 sentences, against 0.2 in prescriber sections. A blind hand check of 100 pairs supports the heading-based alignment for seven of ten heading types. The corpus, its code and its documents are public, and the release keeps texts that show a company copyright mark out of the text tables.

cs.CL · cs.CY

新论文 · v1arXiv:2611.21154

KuroshioVis: Hypothesis-Driven Discovery of Mesoscale Air-Sea Interaction through Coarse-to-Fine Exploration

Yuanzhi Chen, Yiwei Lin, Xi Duan, Qingqing He, Ruiliang Xu, Yucheng Wang, Qiong Zeng

原文摘要

Mesoscale air-sea interaction occurs throughout strong sea surface temperature fronts such as the the Kuroshio Extension, but its strength is highly non-uniform and can be resolved only in kilometre-scale coupled simulations. The response of any single eddy is weak relative to synoptic variability, so evidence supporting a hypothesis of systematic air-sea coupling must be accumulated across many eddies rather than inferred from individual cases. We present KuroshioVis, a hypothesis-driven, coarse-to-fine visual analytics workflow. The coarse stage extracts and aligns mesoscale eddies into composite representations of their air-sea responses, and provides a global view for exploration. The fine stage investigates these responses through three coordinated views: the composite in terface view, the cross-interface flow view, and the evidence matrix view. These coordinated views support air-sea coupling analysis from surface patterns to cross-interface dynamics and quantitative statistical evidence. We demonstrate the workflow using the coupled C1440-LLC2160 simulation of the 2026 SciVis Contest.

cs.HC · cs.GR

版本更新 · v4arXiv:2610.05570

Poisson-GENERIC Neural Operators: Exact Metriplectic Structure in Function Space via Casimir Entropies

Jason Sulskis, Sathya Ravi

原文摘要

Existing thermodynamically consistent neural operators enforce GENERIC degeneracy by projecting the reversible operator, introducing state dependence and generally forfeiting the Jacobi identity. We instead use the compatible Lie--Poisson pencil $L=αD+λ(uD+Du)$ for nonlinear transport and constant Poisson multipliers otherwise. Augmenting $u$ with a latent entropy density $s$ makes $S=\int s$ a Casimir, so $L\,δS/δz=0$ without projection. The energy combines a fixed mechanical quadratic, a learned gauge-free potential, and a convex internal energy; the friction operator $M=AA^{\!\top}$ enforces $M\,δE/δz=0$ pointwise and supports diffusion and damping through local Onsager fluxes. Skewness, positivity, and both degeneracies hold algebraically for any parameters; the continuum pencil satisfies Jacobi, with discrete residuals at machine precision on alias-free inputs. Closed-form heat and damped-wave friction operators are exactly representable, and a discrete-gradient integrator preserves the first and second laws. Across four PDEs in one and two dimensions and three backbones (FNO, Transolver, CNO), the model wins $62/72$ seed-level comparisons against unconstrained baselines at equal optimizer steps, recovers exact transport and wave symbols, matches heat and Burgers dissipation within $12\%$, and dissipates nothing on advection. At equal wall-clock time, the unconstrained FNO is more accurate over ten frames. Over one hundred frames, in and out of distribution, our model never increases physical energy and conserves it when dissipation vanishes, whereas the unconstrained FNO loses a quarter of conserved energy out of distribution and projected GENERIC-FNO often creates energy. Limitations include late-time heat under-damping and explicit-integrator divergence on waves.

cs.LG

新论文 · v1arXiv:2611.21149

Cap-Aware Boosting for Green Day-Ahead Electricity Price Forecasting

Shubham Pratap

原文摘要

Accurate forecasting of Green Day-Ahead Market (GDAM) electricity prices is essential for market participants in the Indian power sector. We present a LightGBM-based forecasting framework for Indian Energy Exchange GDAM prices, which clear in 96 fifteen-minute intervals per day. The proposed system explicitly models price-cap behavior by separating capped-price events from normal market conditions. A classification head estimates the probability of the regulatory price cap (Rs 10,000/MWh), while a regression head predicts uncapped prices in logarithmic space. Final forecasts are obtained through probabilistic blending of these components and are accompanied by nine quantile forecasts for uncertainty estimation. Evaluated on 257 trading days between 1 January and 14 September 2026, the model achieves a mean absolute error (MAE) of 397.9 Rs/MWh and a mean absolute percentage error (MAPE) of 9.57%, substantially outperforming a Chronos-2 baseline (583.3 MAE) and a persistence baseline using the previous day's profile (589.1 MAE). We further introduce leak-safe variants that incorporate real-time market signals, unmatched sell volumes, and solar generation proxies. In a recent out-of-sample evaluation, the quantile-clock variant achieved the best short-term performance with an MAE of 789 Rs/MWh during 25-28 October, while the enhanced capped-price model achieved an MAE of 716 Rs/MWh across 21 evaluation days through 16 November. These results demonstrate the effectiveness of regime-aware probabilistic forecasting for high-volatility electricity markets.

cs.LG

新论文 · v1arXiv:2611.21146

Model-based neural network for global floodplains analysis

Pierre Audisio, Barbara Belletti, Nelly Pustelnik

原文摘要

We address the super-resolution of multispectral remote sensing imagery for long-term floodplain monitoring, enhancing Landsat acquisitions towards Sentinel-2 resolution. Exploiting the period during which both missions observed the same areas, we build DLhasa-L2S , to our knowledge the first paired dataset dedicated to multispectral super-resolution for river analysis. It covers the Lhasa River (Tibet), whose strong anthropogenic changes and diverse hydromorphological profiles make it an ideal study area. We developed two model-based neural networks: a forward-backward (FB) plug-and-play scheme and an unfolded FB scheme. We compared them against learned neural networks baselines (SwinIR, RDN) on synthetic data, on real scenes, and through a statistical test over the whole real paired Landsat-Sentinel dataset. The unfolded network achieves the highest PSNR and SSIM on the real RGB components with fast inference, offering the best trade-off between accuracy and computational cost. For NDWI-based water segmentation, its performance shows no statistically significant difference from that of the best methods.

cs.CV

版本更新 · v2arXiv:2605.01337

Comparison of OpenAlex and Scopus coverage of German institutions' publications in top-tier journals

Andrey Lovakov, Ivan Sterligov

原文摘要

OpenAlex has recently emerged as a leading alternative to proprietary bibliometric sources. However, concerns remain regarding the quality of its metadata, especially the institutional profiles which are crucial for evaluating organizations. This study assesses the quality of affiliation data in OpenAlex using German research institutions. Publications from top-tier journals were analyzed and institutional publication counts in OpenAlex were systematically compared with counts in Scopus. The results show that OpenAlex generally contains more publications at the journal level, reflecting its broader coverage. However, institutional publication counts in OpenAlex are consistently lower; it contained 6.8% fewer institution-paper links in the analyzed sample. Aggregate institutional counts nevertheless showed very high rank correlation and concordance with Scopus. Discrepancies varied substantially across fields, publishers and journals. Detailed investigation identified missing affiliation information, differences in institutional boundaries, and divergent year and journal assignments, while errors present in both databases. These findings suggest that OpenAlex is suitable for comparative institutional analyses in academic research but requires further improvement in affiliation metadata before it can be used for evaluation contexts that rely on absolute publication counts.

cs.DL

新论文 · v1arXiv:2611.21143

WinDEX: Metric 3D Hand Motion Reconstruction in the Wild for Dexterous Teleoperation and Policy Learning

Hongyu Ma, Hairong Qu, Shiqi Zhao, Yongsong Yang, Kai Shi, Jiahao Ai, Xin You, Peng Yin

原文摘要

Human manipulation data are widely used to train robots for human-like dexterity. However, low-cost, large-scale collection in the wild remains challenging. Existing methods remain limited in geometrically grounded metric-scale reconstruction, real-world generalization and inference speed. We present WinDEX, integrating (1) geometric-temporal optimization for metric 3D hand annotation from in-the-wild egocentric stereo videos, combined with visual-inertial SLAM for world-frame reconstruction; (2) WinDEX-Net, an end-to-end, fisheye-native network combining calibrated stereo, local detail and temporal context, with a data-model flywheel of model-assisted annotation and retraining; (3) distilled models for offline reconstruction and causal streaming, supporting reinforcement learning and teleoperation; and (4) low-cost CyberCap stereo-inertial capture, forming the WinDEX-Hand dataset with 1,000 hours of training annotations and held-out human-annotated in-the-wild 2D and independent optical motion-capture evaluation sets. Offline reconstruction runs at 43.61 FPS and causal streaming at 171 FPS. On held-out data, offline reconstruction achieves state-of-the-art accuracy: 14.5 pixels in-the-wild reprojection error and, with visual-inertial odometry, 37.29 mm world-frame joint error. We achieve 7.6 and 3.6 times the task success rates of reinforcement learning policies trained with WiLoR and HaWoR motion references, respectively, and demonstrate real-time arm-hand teleoperation. Project website: https://fruitrabbit.github.io/WinDEX/.

cs.RO

版本更新 · v2arXiv:2607.12226

Flex: A Foundational Proof-Producing Horn Solver

Jam Kabeer Ali Khan, Petros Markopoulos, Niki Vazou, Nico Lehmann, Ranjit Jhala

原文摘要

Constrained Horn Clauses (CHCs) are a versatile intermediate representation for program verification, but leading CHC solvers rely on SMT-based reasoning over supported background theories. Specifications involving user-defined recursive functions or bespoke operators can require additional semantic encodings and reasoning beyond these theories. We observe that CHCs are existentially quantified propositions in a foundational logic, and so can be proved within an interactive theorem prover, where specifications can use native definitions, and users can complete proofs when automation stalls. However, in a foundational setting, every Horn variable solution must be justified by a kernel-checked proof. We present Flex, a proof-producing CHC solver implemented as a Lean metaprogram, that combines predicate synthesis with proof construction. Flex comprises a suite of tactics: Zap, which synthesizes strongest predicates for acyclic Horn variables and exploits their structure to construct elimination proofs without any proof search; Fix, which uses predicate abstraction over templates built from arbitrary Lean terms, discovered by scraping the CHCs and associated definitions, to infer inductive invariants for cyclic Horn variables; and a certifying pipeline that keeps external SMT-based solvers like Spacer and Eldarica outside the trusted computing base. We use Flex to verify the functional correctness of sorting algorithms, hash tables, ring buffers from the Tock OS kernel, and Rust std library challenges. We show that Zap constructs elimination proofs 28.2x and 44.6x faster than grind and aesop, respectively. With Zap and Fix, Flex automatically proves 98.0% of the refinement-typechecker CHCs in our evaluation.

cs.PL · cs.LO · cs.SE

新论文 · v1arXiv:2611.21137

Scribble Once, Segment the Sequence: Bidirectional Temporal Consistency for Weakly Supervised Segmentation in Echocardiography Sequences

Vincent Li Weixin, Chen Jingkun, Vicente Grau

原文摘要

Temporal medical image segmentation relies on pixel-wise annotations, which are prohibitively expensive for long image sequences. Existing weakly-supervised methods reduce the annotation burden per frame but still require repeated annotations across the sequence. Furthermore, sparse supervision can result in temporally unstable representations. Addressing both challenges requires a model that can leverage strong spatial priors from large-scale pretraining while minimizing reliance on dense annotations. Frozen 2D vision foundation models provide a promising basis for this by offering transferable spatial representations with low trainable parameters. However, their lack of explicit temporal modeling limits their ability to propagate reliable information across anatomical sequences. To address this, we present a weakly-supervised framework for temporal segmentation from a single scribble-annotated frame, with all remaining frames left unlabeled. The framework combines parameter-efficient adaptation of DINOv3 encoders present in the mean-teacher architecture. Furthermore, a set of complementary losses, including a temporal loss helps generate reliable segmentation priors for the unlabeled frames. This formulation decouples annotation costs from sequence length while retaining the transferable spatial knowledge of the pretrained encoder. Extensive experiments on the publicly available CAMUS and CardiacNet datasets demonstrate that our approach outperforms other scribble-supervised methods. These results show that accurate and temporally consistent segmentation can be achieved from extremely sparse supervision, substantially reducing annotation requirements for temporal echocardiographic segmentation and providing a practical framework for scalable clinical deployment.

eess.IV · cs.CV

版本更新 · v2arXiv:2605.26322

OmniToM: Benchmarking Theory of Mind in LLMs via Explicit Belief Modeling

Adam Bawatneh, Sagar Sapkota, Amrit Singh Bedi, Santu Karmaker, Mubarak Shah

原文摘要

Theory of Mind (ToM), the ability to infer others' knowledge, intentions, and emotions, is commonly evaluated in large language models (LLMs) using end-point question answering, where performance is judged solely by the final answer to a social reasoning query. This paradigm obscures whether the model actually constructs the underlying mental-state representations required for robust reasoning, particularly in scenarios involving divergent, evolving, or mistaken beliefs. In order to address this research gap, we introduce OmniToM, a benchmark that directly evaluates these representations by requiring explicit modeling of belief structures for all relevant actors within a narrative. These structures are composed of belief propositions: minimal statements of what an actor takes to be true about the world or another actor's mental state, allowing knowledge, intentions, emotions, and false beliefs to be analyzed in a common format. Models are evaluated in two stages: Stage 1: Belief Extraction, which extracts from the story the beliefs relevant to its social dynamics, and Stage 2: Belief Labeling, which assigns each belief a seven-dimensional schema label covering recursive order, truth status, knowledge access, explicitness, content type, mental source, and context. Built from 895 stories from the existing ToMBench story corpus and augmented with 22,343 labeled belief propositions, OmniToM uses a human-calibrated LLM-assisted annotation pipeline. Across diverse models in zero-shot evaluation, OmniToM reveals an actor-specific belief-tracking bottleneck: current LLMs struggle with the knowledge-access and representational decisions required to transform narrative facts into actors' beliefs and shared mental states.

cs.AI

版本更新 · v2arXiv:2603.18657

Enhancing Multi-Corpus Training in SSL-Based Anti-Spoofing Models: Domain-Invariant Feature Extraction

Anh-Tuan Dao, Driss Matrouf, Mickael Rouvier, Nicholas Evans

原文摘要

The performance of speech spoofing detection often varies across different training and evaluation corpora. Leveraging multiple corpora typically enhances robustness and performance in fields like speaker recognition and speech recognition. However, our spoofing detection experiments show that multi-corpus training does not consistently improve performance and may even degrade it. We hypothesize that dataset-specific biases impair generalization, leading to performance instability. To address this, we propose an Invariant Domain Feature Extraction (IDFE) framework, employing multi-task learning and a gradient reversal layer to minimize corpus-specific information in learned embeddings. The IDFE framework reduces the average equal error rate by 20% compared to the baseline, assessed across four varied datasets.

cs.LG

新论文 · v1arXiv:2611.21132

Region-VAD: Video Anomaly Localization and Detection via Region Selection

Giacomo D'Amicantonio

原文摘要

Anomalies are local events, yet weakly supervised video anomaly detection scores them globally. Most methods pool an entire frame into a single vector, rank frames by that vector, and derive a location afterwards as a saliency map of the score. Full-frame pooling mixes the anomaly with its background, so a detector can reach competitive benchmark scores from scene context without being able to show where in the scene the anomaly occurs. We introduce Region-VAD, which first selects a region and then detects the anomaly from that region alone. Region-VAD selects an explicit rectangle over the tokens of a frozen foundation model and scores anomalousness only from the tokens inside the box, through attention pooling followed by a standard temporal head. The rectangle is not learned, it is the box in which the frozen tokens change most between consecutive segments, in all four feature projections of the encoder, and it is found by exhaustive search over the rectangles of the token grid. Only the pooling and the temporal head are trained, from video-level labels without boxes, masks or frame-level annotations, and no trained component reads a token outside the selected region. Experiments on five datasets, with one localizer configuration shared by all datasets and backbones, show that Region-VAD achieves state-of-the-art anomaly localization on UCF-Crime and MSAD at competitive detection performance.

cs.CV

新论文 · v1arXiv:2611.21129

Linglong Agent Service: A Runtime for Centralized Agent Reasoning and Distributed Execution

Hua Zhang, Min Jiang

原文摘要

LLM agents can use remote tools and cloud sandboxes, but integrating reasoning, session state, and heterogeneous execution sites remains a runtime design problem. We present Linglong Agent Service (agentsvc), a self-hosted HTTP service and embeddable Python runtime that treats execution site as a configurable dimension alongside model, tools, and context. A parent service retains inference orchestration, tool descriptions, and conversation history; each inference request selects one child environment for file, shell, terminal, and MCP operations. Direct HTTP and child-initiated WebSocket tunnels carry the same application interfaces, including portable request context and centrally managed updates. The normal core path requires no third-party Python packages, with an optional Windows pseudo-terminal dependency. We further present online_debugger as the process-scale special case of execution-site separation: an instrumented program submits exception evidence, executes session-scoped inspection tools inside the incident process, and resumes the same diagnostic session. Its default prompt requires failing-then-passing reproduction evidence for source changes, while runtime patch activation and failed-call recovery remain unimplemented. We explain the design through constrained expected-utility reasoning, compare concrete mechanisms with Hermes Agent, Pi, Codex, and Claude Code, and report implementation and regression-test evidence. The current evidence establishes executable contracts, not performance advantages, repair accuracy, or production recovery. The source code is open source under the MIT license at https://github.com/mz24cn/agents.

cs.RO · cs.AI

版本更新 · v4arXiv:2506.16552

Revela: Dense Retriever Learning via Language Modeling

Fengyu Cai, Tong Chen, Xinran Zhao, Sihao Chen, Hongming Zhang, Sherry Tongshuang Wu, Iryna Gurevych, Heinz Koeppl

原文摘要

Dense retrievers play a vital role in accessing external and specialized knowledge to augment language models (LMs). Training dense retrievers typically requires annotated query-document pairs, which are costly to create and scarce in specialized domains (e.g., code) or in complex settings (e.g., requiring reasoning). These practical challenges have sparked growing interest in self-supervised retriever learning. Since LMs are trained to capture token-level dependencies through a self-supervised learning objective (i.e., next token prediction), we can analogously cast retrieval as learning dependencies among chunks of tokens. This analogy naturally leads to the question: How can we adapt self-supervised learning objectives in the spirit of language modeling to train retrievers? To answer this question, we introduce Revela, a unified and scalable training framework for self-supervised retriever learning via language modeling. Revela models semantic dependencies among documents by conditioning next token prediction on local and cross-document context through an in-batch attention mechanism. This attention is weighted by retriever-computed similarity scores, enabling the retriever to be optimized as part of language modeling. We evaluate Revela on domain-specific (CoIR), reasoning-intensive (BRIGHT), and general-domain (BEIR) benchmarks across various retriever backbones. Without annotated or synthetic query-document pairs, Revela surpasses larger supervised models and proprietary APIs on both CoIR and BRIGHT. It achieves BEIR's unsupervised SoTA with ~1000x less training data and 10x less compute. Performance increases with batch size and model size, highlighting Revela's scalability and its promise for self-supervised retriever learning.

cs.IR · cs.CL

版本更新 · v2arXiv:2611.17019

Temporal Regularized Learning: Self-Supervised Learning Local in Space and Time

Davide Wiest

原文摘要

Self-supervised objectives such as VICReg learn representations by making related inputs similar, while variance and covariance penalties prevent collapse. These penalties are statistics over a batch of samples, so layer-local variants, too, are typically local in space but not in time. We introduce Temporal Regularized Learning (TeReL), which optimizes a VICReg-style objective one sample at a time with purely local updates: each update uses only the current activity of its own and the preceding layer and a fixed-size running state. Temporally adjacent inputs serve as positive pairs, which makes the objective a soft form of Slow Feature Analysis (SFA). Running estimates of each neuron's mean and variance, together with one lateral matrix per layer, replace the batch statistics; held fixed during differentiation, they turn the gradient into a Hebbian feedforward update and an anti-Hebbian lateral update. We show that a two-layer network trained with batch size one on an MNIST stream of class-homogeneous chunks reaches 96.48% linear-probe accuracy, above random features (95.09%) and classical SFA, and that the temporal order and each loss term contribute to this result. With both locality constraints relaxed, the same objective matches supervised backpropagation. TeReL thus shows that the collapse-preventing statistics of modern self-supervised learning can be held as running state at each neuron, which brings these objectives closer to what brains and neuromorphic hardware can implement.

cs.LG · stat.ML

版本更新 · v2arXiv:2508.16569

A Disease-Centric Vision-Language Foundation Model for Precision Oncology in Kidney Cancer

Yuhui Tao, Zhongwei Zhao, Zilong Wang, Xufang Luo, Feng Chen, Kang Wang, Chuanfu Wu, Xue Zhang 等 20 位作者

原文摘要

The non-invasive assessment of renal masses remains a critical challenge in urologic oncology, where diagnostic uncertainty frequently causes overtreatment. Here, we develop RenalCLIP, a vision-language foundation model for precision oncology in kidney cancer. Utilizing 27,866 computed tomography scans from 8,809 patients across diverse multi-center cohorts, we employ a two-stage pre-training strategy to align domain-specific visual and textual representations. RenalCLIP achieves enhanced performance and generalizability across ten core clinical tasks, spanning anatomical assessment, diagnostic classification, and survival prediction, significantly outperforming state-of-the-art general-purpose foundation models. Furthermore, RenalCLIP demonstrates strong data efficiency in diagnostic classification, achieving peak baseline performance using only 20% of the training data. The model also exhibits robust zero-shot diagnostic capabilities, effective image-text retrieval, and high-quality medical report generation. Our findings establish RenalCLIP as a powerful, generalizable tool to enhance diagnostic precision, refine prognostic stratification, and personalize the management of renal masses.

eess.IV · cs.AI · cs.CV

新论文 · v1arXiv:2611.21119

The Limits of Closed-Form Couplings for Few-Step Generation

Tobias Höppe, Stefan Bauer, Qiang Liu, Andrea Dittadi, Kirill Neklyudov

原文摘要

Few-step generators via flow models require complicated technical solutions such as a pretrained flow model or a consistency loss. While rectified flows have gained popularity for their success in few-step sampling, they pay a cost of needing a pre-trained model and sampling a large number of couplings per dataset. Yet for a finite dataset and a Gaussian probability path, the probability-flow vector field of the empirical distribution is available in closed form, so one could use this field as the teacher and skip pretraining altogether. We demonstrate that, surprisingly, this approach is impossible for problems of practical relevance. In particular, the analytic solution of the diffusion process in high dimensions becomes almost a linear transformation. The integration returns the noise vector drawn at initialisation, plus a drift that is empirically a faint copy of the data point, almost independent of the noise. The coupling therefore differs from an independent pairing only through this drift. As a result, training a model on these couplings is almost equivalent to training the model on the independent draws of the noise and the data, which does not provide a useful teacher. We highlight that this is the direct consequence of the high-dimensionality of the data, and that for low enough dimensions or large enough datasets, analytic couplings become more efficient teachers. We quantify this effect by the fraction of endpoint energy carried by the drift, which scales as $\frac{\log N}{d}$ for a dataset of size $N$ in a $d$-dimensional space. Finally, we provide an empirical study confirming all the theoretical results presented in the paper.

stat.ML · cs.LG

版本更新 · v2arXiv:2602.20805

Assessing the Impact of Speaker Identity in Speech Spoofing Detection

Anh-Tuan Dao, Driss Matrouf, Nicholas Evans

原文摘要

Spoofing detection systems are typically trained using diverse recordings from multiple speakers, often assuming that the resulting embeddings are independent of speaker identity. However, this assumption remains unverified. In this paper, we investigate the impact of speaker information on spoofing detection systems. We propose two approaches within our Speaker-Invariant Multi-Task framework, one that models speaker identity within the embeddings and another that removes it. SInMT integrates multi-task learning for joint speaker recognition and spoofing detection, incorporating a gradient reversal layer. Evaluated using four datasets, our speaker-invariant model reduces the average equal error rate by 17% compared to the baseline, with up to 48% reduction for the most challenging attacks (e.g., A11).

cs.SD · cs.LG

版本更新 · v4arXiv:2605.12567

Coherence-Guided Pyramid Self-Contrastive Disentanglement for Adaptive IQ Signal Denoising in Synthetic Aperture Ultrasound

Jiajing Zhang, Bingze Dai, Xi Zhang, Yue Xu, Wei-Ning Lee

原文摘要

Synthetic aperture ultrasound (SAU) groups neighboring elements in an array into sub-apertures and sequentially transmits them to achieve two-way focusing by compounding all received signals. However, the inherent electronic noise, side lobes, and speckle noise still complicate clinical interpretation of SAU images. Previous denoising methods either rely on explicit noise assumptions or are pretrained in a limited image domain, thus facing inevitable domain shift under complex \textit{in vivo} noise conditions. This study proposes an adaptive IQ signal denoising method to suppress aperture-dependent degradation before B-mode image formation. We model sub-aperture events in each SAU acquisition as multiple noisy measurements and extend cross-aperture coherence as a contrastive learning critic in a Pyramid Self-Contrastive Disentanglement (PSCD) framework. PSCD encodes the raw IQ tensor into separated coherent and incoherent latent spaces, then reconstructs clean IQ signals exclusively from the former to retain coherent tissue information while suppressing incoherent noise. An Aperture-to-Aperture proxy task self-supervises the reconstruction in dual domains. The whole framework can be optimized at test time using single-shot data without pretraining or clean labels. Simulations including noise levels from 0 to 30 dB demonstrated improvements of 69.3$\%$ in SNR and 34.4$\%$ in CNR. \textit{In vivo} experiments, including six echocardiographic views and abdominal images, showed 84.8$\%$ SNR and 25.7$\%$ CNR gains using only two sub-apertures. Our PSCD also achieved higher image resolution and boundary sharpness, while preserving high-frequency details. These results support adaptive IQ signal denoising leveraging cross-aperture measurement redundancy, enabling precise anatomical visualization without domain shift and pretraining costs.

cs.CV · cs.AI

新论文 · v1arXiv:2611.21116

Out-of-training properties in frozen interatomic potential embeddings are decodable and ordered by physical scale

Natalija Stepurko, Joonatan E. M. Laulainen

原文摘要

Machine-learned interatomic potentials (MLIPs) are trained on energies, forces and stresses, yet their frozen representations can predict other materials properties. We probe ORB-v3 and UMA-S across layers, read-outs and equivariant channels for nine properties of 154,875 Materials Project crystals. Both models outperform composition-only descriptors on all targets, including five outside their training objectives; errors in physical units help assess potential screening applications. Optimal depth varies by property, consistent with a physical-scale ordering. Random-weight controls reveal strongly property-dependent benefits of pretraining. UMA-S's rotation-invariant angular channels improve crystal-system classification, while higher scores from its full tensor disappear under rotation; matched-model controls do not isolate equivariance as the explanation. Accuracy varies across different probes, but nearest-neighbour prediction often approaches XGBoost performance, even though materials do not form distinct clusters according to their properties. We also test whether a linear predictor trained on one model's embeddings can predict properties using the other model's embeddings after a learned linear transformation. Predictive accuracy is largely preserved, despite differences in which atoms the two models place close together. These results guide the reuse of frozen MLIP representations for materials prediction.

physics.chem-ph · cond-mat.mtrl-sci · cs.LG

新论文 · v1arXiv:2611.21115

ReuseMem: One Memory, Multiple Goals in Robot Manipulation

Yoonseok Kang, Dayeon Jung, Seunghoon Jeon, Wonjib Kim, Dongchul Park

原文摘要

Robots that work alongside people should be able to use what they have seen when new requests arise. The same experience may later support identifying an object or returning it to its original location, although these uses were unknown during observation. ReuseMem learns to use one visual memory for multiple later manipulation goals. It records visual features independently of future goals, retrieves complementary evidence for each request, and combines it with the current scene through a lightweight action interface. Retrieval is supervised by how well candidate evidence supports expert-action prediction across different current states, linking memory selection to its use in manipulation. Repeated visual features are stored once, and memory computations that remain unchanged during action generation are reused. Experiments on four simulation tasks and two physical scenarios demonstrate reuse of a common recording across different requests. With matched action adaptation, ReuseMem improves joint goal completion on the benchmark tasks by 24.5 percentage points over uniform memory. These findings show that one memory can support multiple later manipulation goals as the scene changes.

cs.RO

新论文 · v1arXiv:2611.21114

VAE Rhythm: Temporal Decoding Traces for AI-Generated Video Detection

Ke Sun, Shuwei Li, Haiyang Mei, Mike Zheng Shou

原文摘要

Visible artifacts are an unreliable cue for detecting generated videos, because newer generators learn to remove them. We instead detect a trace left by the generation pipeline itself. Most video generators decode latents with a 3D causal VAE that groups consecutive frames in time, so the fidelity of generated frames rises and falls periodically with their position in a group. We call this periodic trace VAE Rhythm. Real videos are captured frame by frame and do not have it. To reveal it without training, we propose a phase scan. The scan reconstructs a clip with an off-the-shelf VAE from successive start offsets and measures the periodicity of the reconstruction error across these offsets. We find VAE Rhythm in most of the 39 variants of 31 generators we test, including the closed-source Sora, Kling 3.0 and Seedance 2.0. We use the rhythm as a zero-shot score, as a plug-in for existing detectors and as a data engine that pairs real videos with their VAE reconstructions to train our detector CaDet. On DF26, the training-free rhythm score outperforms all 15 retrained baselines, and as a plug-in it improves all 19 tested detectors by 6.7% to 23.7% AUROC. CaDet achieves the best AUROC on AIGVDBench, DF26 and three 2026 closed-source models, and its variant trained without any generated video still exceeds 90% on AIGVDBench.

cs.CV

新论文 · v1arXiv:2611.21111

Searching for Autoregressive Neural Architectures from Fundamental Operations

Thomas Woehrle, Linus Ericsson, Henry Gouk, Elliot J. Crowley

原文摘要

Expressive Neural Architecture Search (NAS) search spaces are desirable as they do not restrict the space of potential architectures to what is already established. However, few search spaces are expressive enough to represent a broad range of architecture families, and those that are have focused primarily on image classification problems. The task of autoregressive sequence modelling has become ubiquitous in modern machine learning, but no existing search space supports the construction of and search over sequence models from fundamental operations, instead imposing strong inductive biases on the structure of potential architectures. This work describes how the expressive einspace search space can be extended to yield a search space capable of representing autoregressive models using fundamental operations. We perform a series of experiments showing that the novel einspaceAR search space can find capable autoregressive sequence models from scratch--i.e., without imposing strong inductive biases. We believe that our einspaceAR search space will be a useful part of the NAS toolbox and that the techniques described in this work can serve as a blueprint for the integration of autoregressive modelling in other expressive search spaces.

cs.LG · cs.AI