AI conference papers

AI conference paper library

Search papers from 30+ top conferences, including NeurIPS, ICLR, ICML, CVPR, ACL and AAAI. Filter by keyword, then open any paper for its abstract, authors and links.

arXiv

MAGE-Vein: Multi-Instance Age and Gender Estimation from Finger Vein Images

Katsuki Tanaka, Koichi Ito, Takafumi Aoki, Masakazu Fujio, Yosuke Kaga, Kanade Oshima, Kenta Takahashi

Age estimation from finger vein images has been widely considered impractical due to severe demographic biases in public datasets and physiological confounding factors like gender. To overcome these limitations, we propose MAGE-Vein, a novel multi-instance, multi-task learning framework. Our approach extracts robust structural aging signs by employing a hybrid feature-level fusion of three fingers, effectively suppressing local imaging noise. Furthermore, simultaneous optimization of gender classification conditions the network to effectively eliminate gender-specific vascular variations. Evaluated on a demographically balanced dataset of 402 subjects, MAGE-Vein achieves a mean absolute error of 6.12 years and a correlation of 0.880. Our results not only overturn the conventional consensus regarding the limitations of the finger vein modality but also demonstrate that previous estimation failures were primarily artifacts of biased public datasets. Our code is available at https://github.com/gsisaoki/MAGE-Vein.

arXiv

Engine-Native Editable 3D World Reconstruction with Objects and Lighting

Junhao Chen, Xinghao Chen, Henghaofan Zhang, Zihao Qiao, Saining Zhang, Yongzhi Li, Ruqi Huang, Sisi Li, Yimin Sheng, Jianyi Zhu, Hao Zhao

Editable 3D scene creation requires object instances and lights that can be inspected, moved, and imported into standard engines, yet existing single-image methods largely stop at room-scale geometry, baked/global illumination, or text-driven generation. We introduce Lumera (Light-aware Unified Engine-native Reconstruction and Assembly), a benchmark and reference pipeline for engine-native, light-aware 3D scene parsing from a single image. Lumera-2K is built from 2,513 UE5 projects and provides 3.73M components, 63M object instances, 102.6K engine-native parametric lights, and 95.1K camera views. On this data, Lumera-Box and Lumera-Light adapt VLM to parse object boxes and parametric light tuples (x,y,z,r,g,b,I), which are assembled with per-object mesh reconstruction, HDR environment estimation, and a bounded agentic refinement loop. In a sanitized box benchmark against DetAny3D, SpatialLM, N3D-VLM, and WildDet3D, Lumera-Box obtains the strongest overall detection, geometry, semantic, and layout scores (merged mAP 0.1141, IoU-B 0.2472, F-score 0.2762), while WildDet3D remains stronger on anchor recall. For lights, Lumera-Light recovers almost all non-empty scenes (recall 0.998) but remains limited at individual-light localization (F1 0.209 at 0.5 m); matched lights have median position error 0.261 m, median ΔE2000 4.59, and intensity Pearson r=0.628. These results establish parametric lights as a measurable editable-scene target and expose remaining bottlenecks in relation structure, light recall/intensity, and cross-engine generalization.

arXiv

WhereEdit: Mask-aware Local Latent Editing for One-Step Image Editing

Ming Hu, Mingyu Dou, Jianfu Yin, Miaomiao Zhang, Cong Hu, Yao Wang, Bingliang Hu, Quan Wang

Recent one-step text-to-image (T2I) models enable efficient image synthesis and provide new opportunities for real-time image editing. However, existing one-step editing methods primarily rely on text conditioning for semantic transformation, lacking explicit spatial control over \textit{where} to edit. More importantly, even when spatial constraints are introduced, these methods often struggle to achieve strong and stable semantic modifications within the target regions. In this work, we revisit one-step image editing from a spatially controlled perspective and identify two key challenges: discovering editable regions and achieving effective localized semantic transformation. We reveal that existing methods perform global semantic transport, which limits high-intensity local editing under the one-step setting. To address this issue, we propose \textbf{WhereEdit}, a framework that reformulates one-step editing as localized adaptive editing. WhereEdit automatically identifies semantically relevant regions from internal model features and applies adaptive local modulation to enhance target-region editing while preserving non-target areas and structural consistency. Experiments on the PIE-Bench benchmark demonstrate that WhereEdit consistently outperforms existing one-step image editing methods, achieving superior editing quality while maintaining the efficiency of one-step generation. Additional experiments with region-level supervision further highlight the importance of explicit spatial reasoning for high-quality one-step image editing.

arXiv

Webly Supervised Multi-Label Recognition: Evaluation Benchmark and Dual-Branch Multi-Label Contrastive Learning

Zhihua Xu, Zhijing Yang, Yufeng Yang, Tianshui Chen

Training deep learning models with freely available web images can reduce their dependence on costly manual annotations. Although webly supervised learning has been widely studied for single-label recognition, its multi-label counterpart remains underexplored, partly due to the lack of unified benchmarks and fair comparison protocols. To address this gap, we construct a benchmark for webly supervised multi-label recognition (WS-MLR), including Web-COCO and Web-Pascal, and re-implement representative baselines under a unified setting. The two datasets cover the same 80 and 20 categories as MS-COCO and Pascal VOC, respectively, and contain about 300 thousand images retrieved from the Internet using category-word combinations as search keywords. We further propose a Dual-Branch Multi-Label Contrastive Learning (DBMLCL) framework, which learns category-specific instance-level and category-level representations together with their similarities to identify and correct noisy labels. Extensive experiments on the benchmark demonstrate that DBMLCL achieves superior performance compared to representative baselines.

arXiv

ViSTR-Bench: Can MLLMs Reason from Continuous Visual Cues in Dynamic Scenes?

Han Li, Si Liu, Zehao Huang, Dongxin Lyu, Longfei Xu, Jiahui Fu, Daxin Tian, Yuliang Xiu, Naiyan Wang

Multimodal Large Language Models (MLLMs) have achieved remarkable success across diverse expert-level tasks, but they still struggle with fundamental abilities that humans naturally develop through continuous observation of the real world, such as spatial perception and dynamic reasoning. Recent studies have recognized this gap and introduced dedicated benchmarks to evaluate the spatial-temporal capabilities of MLLMs. However, existing benchmarks mostly focus on static scenes or require exact quantitative predictions, leaving intuitive reasoning from temporal cues largely underexplored. In this paper, we introduce the Visual Spatial-Temporal Reasoning Benchmark (ViSTR-Bench), a novel evaluation suite designed to systematically assess whether MLLMs can perform qualitative reasoning from continuous visual cues in dynamic scenes. Guided by the principles of temporal emphasis, reasoning orientation, and qualitative evaluation, ViSTR-Bench establishes a comprehensive four-dimensional evaluations covering Motion Perception, Spatial Relations, Outcome Prediction, and Physical Dynamics. The benchmark comprises 15 distinct subtasks and 1,340 high-quality video question-answer pairs spanning diverse tabletop, indoor, and outdoor scenarios. Extensive evaluations of a broad spectrum of state-of-the-art proprietary, open-source, and specialized spatial MLLMs reveal that, despite their strong general video understanding capabilities, current models still face substantial bottlenecks in complex spatial-temporal reasoning and remain far below human performance.

arXiv

Agentic Designer: Progressive Multi-Agent Collaboration for Structure-Aware Interior Layout Generation

Zhijing Yang, Haocheng Lin, Zhihua Xu, Haojie Li, Keze Wang, Liang Lin, Tianshui Chen

Generating realistic interior furniture layouts that strictly adhere to architectural constraints (e.g., walls, doors, and windows) remains a fundamental challenge in automated spatial design. Existing approaches, primarily based on one-shot generation using diffusion models or Large Language Models (LLMs), lack explicit mechanisms for intermediate geometric constraint verification, often resulting in structural collisions and functionally infeasible arrangements under complex room constraints. To address these challenges, we propose Agentic Designer, a progressive, multi-agent framework that formulates structure-aware interior layout generation as an iterative and constraint-verified decision process. By decomposing layout synthesis into modular stages of proposal, verification, and adjustment, the framework coordinates three specialized agents, a Generator, an Evaluator, and a Refiner, through a Progressive Consensus Mechanism. This mechanism enforces stepwise geometric validation and correction before each placement is committed, thereby preventing error accumulation. To facilitate this structure-aware paradigm and standardize evaluation, we establish InStruct, a comprehensive benchmark that integrates a dataset comprising over 18,000 high-quality, parametrically annotated samples with a novel suite of structure-centric metrics. Extensive quantitative evaluations, qualitative analyses, and user studies show that Agentic Designer significantly outperforms state-of-the-art methods, demonstrating substantial improvements in strict structural adherence and functional design coherence.

arXiv

SubSplat: High-Resolution Pixel-aligned 3DGS via Sub-pixel Gaussian Reparameterization

Jiun Lee, Jaekwang Kim, Sangmin Lee

Pixel-aligned Gaussian splatting enables efficient and generalizable novel-view synthesis. However, high-resolution rendering faces a critical trade-off where increasing input resolution improves detail at the expense of quadratically rising network computational cost. Conversely, maintaining low-resolution inputs stabilizes this cost but results in insufficient Gaussian density and artifacts. To address this, we propose SubSplat, which introduces Sub-pixel Gaussian Reparameterizer(SPGR) to subdivide primary Gaussians into fine-grained primitives, restoring structural density directly from low-resolution features. We further enhance the reparameterization quality through feature aggregation, which effectively captures high-frequency details across multiple views. Experiments on RealEstate10K and ACID demonstrate that SubSplat achieves high-fidelity rendering with superior efficiency. Our results validate that the proposed framework successfully resolves the trade-off between reparameterization fidelity and network computational cost inherent in pixel-aligned Gaussian Splatting.

arXiv

Toward Seasonal Guidelines for Robust Deep-Learning Sentinel-2 Building Detection in Different Area Types

Michał Romaszewski, Kamil Drejer, Katarzyna Kołodziej, Anna Zawadzka, Stanisław Lewiński, Przemysław Głomb, Marek Ruciński, Michal Krupiński, Krzysztof Gryguc Przemysław Sekuła, Szymon Sala

Sentinel-2 imagery offers open access, global coverage, and frequent revisit times, making it attractive for practical building mapping at scale; however, its native 10m resolution makes building vs non-building classification challenging, particularly for small or sub-pixel buildings, and performance can vary with both seasonality and the heterogeneity of built-up environments. This paper introduces a Sentinel-2 building-detection framework designed to systematically quantify these effects and to support more formalised, practice-oriented model selection. We construct a dedicated multi-temporal Sentinel-2 dataset over the Warsaw region and derive binary ground-truth masks by rasterising official Polish topographic database (BDOT10k) building footprints onto the Sentinel-2 pixel grid. Using two established convolutional segmentation backbones (U-Net and DeepLabV3+), we first perform scene-specific fine-tuning to select a robust architecture and identify the best monthly models for L1C and L2A products separately. We then conduct cross-temporal inference by applying each best monthly model to all scenes, enabling an assessment of (i) which months provide favourable training and inference conditions, (ii) how performance transfers between seasons, (iii) the impact of processing level, and (iv) how these effects differ across built-up typologies. Based on these results, we provide practical guidance for routine Sentinel-2 building classification under varying acquisition periods and settlement characteristics.

arXiv

EgoRecovery: Acquiring Failure Recovery Ability Through Human Recovery Demonstration

Zuhao Ge, Yuchen Zhou, Weitao Zhou, Minglei Li, Xinyu Li, Chao Wu, Hanwen Zhao, Haotian Wang, Zuxuan Wu, Xiaosong Jia, Yu-Gang Jiang

Robust embodied robots should be able to recover from failures and retry tasks in order to operate reliably in unstructured and noisy real-world environments. Achieving this capability requires training policies on data that captures recovery behaviors. However, collecting such data through robot teleoperation is difficult to scale, as it is time-consuming to induce diverse failure states, perform corrective actions, and reset the environment. This challenge is further exacerbated by the high diversity of failure modes, which demands substantially more recovery data than success demonstrations. In this work, we show that egocentric human data capturing failure recovery processes provides a scalable alternative. By efficiently arranging task-level failure configurations and recording short recovery segments, human operators can generate more than 10x as much valid recovery data per hour compared to robot teleoperation under our protocol. To address the embodiment gap between human and robot, we propose EgoRecovery, a co-training framework for learning recovery behavior, where human recovery demonstrations are aligned to a compact corrective-intent space shared with robot data, which captures the timing and magnitude of correction. Only a small number of robot recovery demonstrations are required to connect this intent to executable robot actions. At deployment, a learned recovery gate predicts when correction is needed from robot observations and activates the corrective intent only in recovery states. Experiments on real-world recovery tasks show that EgoRecovery improves success from failure starts over robot-only recovery, direct co-training with human recovery data, and direct intent-transfer baselines.

arXiv

ERank in Latent Space as an Image-Complexity and Richness Measure

Maksim Smirnov, Grigory Kononov, Anastasiia Linich, Egor Surkov, Egor Shvetsov

We propose the effective rank (ERank) of the channel covariance of an image's deep feature map as a per-sample, label-free measure of visual richness, computed from a single forward pass through a frozen pretrained encoder. ERank counts how many decorrelated channel directions an image activates, and we characterize its properties, including its behavior under noise. Empirically, ERank orders images from plain to visually rich, correlates with codec bitrate, sharpness, and edge density, and correlates with human complexity annotations on IC9600 with $r = 0.72$. As a data-selection criterion, removing low-ERank samples improves super-resolution and removing high-ERank samples improves OCR, in both pretraining and finetuning, while selection does not help classification, segmentation, or denoising. ERank is thus a cheap richness signal, useful exactly when task difficulty is governed by input richness.

arXiv

Anticipate Before Acting: Future-State-Conditioned Vision-Language Navigation

Lingfeng Zhang, Zhanguang Zhang, Liheng Ma, Tongtong Cao, Yingxue Zhang

End-to-end vision-language navigation (VLN) with causal vision-language models maps instructions and egocentric observations directly to actions, but standard behavior cloning supervises only the next action and does not explicitly encourage the policy state to be predictive of future visual outcomes, limiting long-horizon decision making. A privileged-input diagnostic shows that access to an expert-trajectory future image can substantially improve navigation, indicating that future observations contain rich, actionable cues, though such inputs are unavailable at deployment. Motivated by this signal, we propose Future-State-Conditioned VLN (FSC-VLN), a deployable model that augments a causal policy with a future-query token and uses training-only future-state supervision to distill information from future observations into the policy state. Concretely, during training we align the future-query representation to a frozen visual embedding $Δ$ steps ahead, while inference requires only past and current observations. This design preserves the baseline inference pattern and adds only two learned prefix tokens, implying minimal overhead. On R2R val-unseen, FSC-VLN improves SR/OSR/SPL over a StreamVLN-style baseline under two training-data regimes, with larger gains on long-horizon episodes; ablations further support the dual-query design that separates future and action queries.

arXiv

DriveFace: A Cross-Spectral Through-Glass Face Dataset for On-the-Move Vehicular Border Control

Anjith George, Luis S. Luevano, Alain Komaty, Zeina Al Amine, Vidit Vidit, Sebastien Marcel

The continuous growth in cross-border mobility places increasing pressure on existing border control infrastructures, motivating on-the-move biometric authentication, in which travellers are identified directly inside their vehicles at checkpoints. Face recognition is well-suited to this setting, as it can be acquired passively and at a distance. Its development, however, is hindered by the lack of representative datasets: existing benchmarks are collected in controlled environments and do not capture the challenges inherent to vehicular acquisition, including motion blur, variable illumination, occlusions, and cross-spectral enrollment. To address this gap, we introduce a dataset for on-the-move face recognition in border-control scenarios, comprising NIR vehicle-crossing videos paired with smartphone-based pre-enrollment data. Baseline evaluations with state-of-the-art models show clear performance limitations under these realistic conditions, highlighting the need for dedicated methods to advance the field.

arXiv

Reflecting Process Expertise in Procedural Material Generation

Kunal Gupta, Gaurav Joshi, Yen-Ru Chen, Seemandhar Jain, Ishit Mehta, Manmohan Chandraker

Procedural material creation underpins applications in digital content creation, visual effects, and 3D asset design. Achieving high-quality results requires more than reproducing node graphs -- it demands understanding the process by which experts construct materials. We formulate procedural material generation as retrieval-time process reasoning over expert demonstrations, elevating process to a first-class representation beyond graph-only synthesis. Concretely, we represent expert workflows as process traces: textual records of construction steps, parameters, and design intent. To instantiate this idea, we use a pretrained LLM-based ProcessSynthesizer to synthesize a process trace aligned with a user's intent and a pretrained LLM-based Compiler to ground the process trace into an executable Blender material graph. Because procedural expertise is most naturally conveyed through demonstrations, we leverage tutorial videos as a source of process knowledge and extract textual, LLM-compatible traces using automated video analysis tools. In an expert study with five Blender artists (avg. 7.5 years of experience), materials generated by reflecting expert demonstrations were found to produce workflows requiring fewer edits, and more closely match professional design strategies than methods operating solely on static artifacts. A user study with 150 participants further shows that our approach achieves superior generation and editing performance compared to prior procedural systems. All code, models, and data will be available at https://materialapprentice.github.io

arXiv

Learning to Navigate Efficiently with Only 0.58M Trainable Parameters

Edward Beng Wai Tan, Siew-Kei Lam

Recent progress in visual navigation has largely been driven by scale: end-to-end policies with hundreds of millions of parameters trained on billions of frames or large-scale simulated data. We ask how much of this scale a single task family actually requires, and what structure can substitute for it. We propose a decomposed navigation model in which operations with known closed-form structure, such as projective geometry, occupancy, and coordinate transforms, are computed analytically and serve as interfaces between three small learned modules: an egress predictor that grounds the episode goal as a local subgoal in the current view, a navigation predictor that estimates a goal-conditioned posterior over where trajectories travel, and an endpoint-pinned residual diffusion generator that samples trajectory shapes from this posterior. The system trains only 0.58M out of a total of 22.7M parameters, on 44k frames in under one GPU-hour, yet approaches the performance of state-of-the-art models on navigation tasks across 6060 point-goal episodes and 60 environments, while having 233$\times$ fewer trainable parameters, the lowest collision rate among all evaluated methods, and 10+ Hz inference rate on a Jetson Xavier NX. The decomposition further transfers to no-goal exploration by retraining only the 123k-parameter egress head, and its failure modes under sensor corruption are transparent and analytically correctable.

arXiv

AnchorPrune: Relevance-Anchored Contextual Expansion for Visual Token Pruning

Kyuan Oh, Bumsoo Kim

Large vision-language models incur substantial inference costs because high-resolution inputs introduce thousands of visual tokens, many of which are redundant for a given query. Existing pruning methods often combine query relevance and token diversity, yet these objectives can conflict under aggressive compression: relevance-driven selection may overconcentrate the budget on correlated local evidence, while diversity-driven selection may suppress indispensable tokens or retain distinct but uninformative regions. We introduce AnchorPrune, a training-free framework that first constructs a protected relevance anchor and then expands it with complementary visual context. AnchorPrune adaptively determines the anchor size from the novelty profile of relevance-ranked tokens, preserving a compact set of query-critical evidence, and allocates the remaining budget through importance-weighted novelty to recover informative, non-redundant context relative to the anchor. This ordered design prevents contextual expansion from displacing indispensable query cues while improving overall visual coverage. AnchorPrune is lightweight, architecture-aware, and requires neither retraining nor model modification. Across image and video vision-language models and benchmarks, it consistently improves the accuracy-efficiency trade-off over training-free baselines, particularly under severe compression. On LLaVA-NeXT-7B, AnchorPrune preserves 97.6% of full-token performance using only 160 of 2,880 visual tokens. These results establish relevance-anchored contextual expansion as an effective principle for efficient multimodal inference. Code is available at https://github.com/MULTI-cau/AnchorPrune.

arXiv

When Does High-CFG Diffusion Inversion Fail? A Controlled Study of Prompt--Latent Interactions

Yan Zeng, Yusuke Hosoya, Huyen T. T. Tran, Takayuki Okatani

Text-guided diffusion inversion is central to image editing, where an image is mapped to an initial latent and then edited by replaying the denoising process under a modified prompt. In practice, however, inversion is often performed with a lower classifier-free guidance(CFG) scale than the one used for generation or editing. This mismatch is empirically useful but leaves a basic question unresolved: when a target image is generated by a high-CFG trajectory, when can that trajectory actually be inverted? We study this question in a controlled generation--inversion--reconstruction setting, where the true initial latent and denoising trajectory are known. Using prompts taken from an existing diffusion-editing benchmark, we generate images under high CFG and reconstruct them with fixed-point inversion using the same prompt and guidance setting. The results reveal three types of prompt-level reconstruction behavior: easy prompts that reconstruct for most initial latents, hard prompts that fail for most initial latents, and intermediate prompts whose success depends on the prompt--latent pairing. To analyze the generation side, we define prompt pressure, a step-wise measure of how strongly CFG moves the denoising update away from the unconditional trajectory. Total pressure correlates with reconstruction quality and separates easy from hard prompts, but it does not explain the success or failure of intermediate prompt--latent pairs. Text-side analyses further show that the main visual subject and wording can change inversion difficulty. Finally, we evaluate a compact trajectory-consistency intervention that relaxes guidance only at locally unstable inverse steps. This diagnostic check improves reconstruction and Prompt-to-Prompt editing in our controlled setting, supporting the view that high-CFG inversion failure requires local, trajectory-aware analysis.

arXiv

On Pairwise Quantile Regression - Statistical Guarantees and Applications

Romain Thérézien, Stephan Clémençon, Fantin Girard, Hamza El-Abdouni

Quantile regression provides a powerful tool for summarizing the conditional distribution of a real-valued random variable (r.v.) of interest $Y$ as a function of covariates $Z$ in cases where it shows a large dispersion with high probability, going beyond the situation where standard least square regression is informative/predictive. This article aims to extend this methodology to the pairwise setting, where the variable to be explained is a similarity score between two independent observations (e.g., pixelated ID photos used as input to biometric systems), and the explanatory variables consist of the pair of covariates attached to these observations, such as age or hair color. We establish theoretical guarantees for solutions of this statistical learning problem, considered here as empirical minimizers of a pairwise version of the pinball loss. Leveraging sharp concentration results for $U$-processes, we prove generalization bounds and identify mild conditions under which fast learning rates can be achieved. Confirming the probabilistic analysis, experiments based on simulation data also provide solid empirical evidence of the validity of the methodology promoted here for pairwise quantile regression. Finally, its usefulness from an application perspective is demonstrated by a detailed study aimed at analyzing errors in similarity scoring for facial recognition.

arXiv

MV-GEL: Language-Driven Multi-View Geometric Entity Localization on Meshes

Kartik Bali, Roland Aydin

Identifying and grounding precise geometric entities, such as edges, planar regions, and curved surfaces within 3D objects, is foundational to computer-aided design (CAD), robotic manipulation, and scientific simulation. Although modern Vision Language Models (VLMs) have advanced referring segmentation (RIS) in the image domain, extending such language-driven localization to structured 3D geometry is substantially harder. The 3D object appearance is highly sensitive to viewpoints; a single perspective may render a target entity clearly observable, while another may suffer from severe occlusion or foreshortening. In this work, we attempt to solve these challenges with MV-GEL (Multi-View Geometric Entity Localization), a framework for localizing fine-grained geometric entities on polygon meshes from natural language queries. Our key insight is that reliable CAD entity (i.e., faces, edges or solids) localization depends on selecting views that make the queried entity maximally interpretable. We introduce GELviews, a prompt-conditioned ranking module that prioritizes viewpoints based on language prompted observability of geometric CAD entities. Selected views are processed by a VLM-based reasoning segmentation backbone, and predicted masks are lifted to the corresponding meshes via geometry-aware ray casting. Our framework is completely CAD agnostic and relies only on 3D meshes. Experiments show up to a 1.7X improvement in face-level IoU and over 4.5X gains in edge-level F1 compared to vanilla baselines, substantially outperforming CLIP-based and random view sampling, particularly for thin and view-sensitive structures. The dataset, code and trained checkpoints are available at https://github.com/kbali1297/MV-GEL.

arXiv

S-Agent: Spatial Tool-Use Elicits Reasoning for Spatial Intelligence

Yalun Dai, Hao Li, Shulin Tian, Runmao Yao, Yuhao Dong, Fangzhou Hong, Zhaoxi Chen, Fangfu Liu, Tao Wang, Kim-Hui Yap, Ziwei Liu

Real-world spatial intelligence requires reasoning over a continuous and evolving 3D world, yet existing VLMs and tool-augmented agents largely remain tied to static, stateless inference from isolated visual observations. We introduce \textbf{\textsc{S-Agent}}, a spatial tool-use agentic paradigm for understanding and reasoning over continuous multi-view images and videos. By formulating spatial reasoning as spatio-temporal evidence accumulation rather than isolated frame-level prediction, \textsc{S-Agent} reshapes spatial perception into scene-centric understanding beyond frame-centric recognition. Specifically, \textsc{S-Agent} casts the VLM as a semantic planner that decides what evidence is needed, while a hierarchy of spatial tools and experts grounds objects in 2D, lifts them into 3D geometric evidence, and aggregates this evidence into high-level spatial knowledge (\textit{e.g.}, counting, measurement, orientation, and relative position). Additionally, a temporal memory mechanism, including Scene Memory for maintaining the evolving scene state and Agent Memory for accumulating reasoning context, enables evidence integration across frames and reasoning steps. Comprehensive experiments on multi-view and video spatial reasoning benchmarks show that \textsc{S-Agent} consistently improves both open-source and closed-source VLMs in a training-free manner. Beyond inference-time augmentation, supervised fine-tuning (SFT) on \textsc{S-Agent}-generated spatial trajectories \textsc{S-300K} yields \textsc{S-Agent-8B}, a compact spatial agent that significantly surpasses similar-scale baselines (e.g., Qwen3-VL-8B) and performs comparably to advanced closed-source models (e.g., GPT-5.4 and Gemini 3).

arXiv

ETC: Extreme Token Compression via Task-aware Visual Information Distillation in VLMs

Yiling Gao, Hongchen Wei, Zhenzhong Chen

In Vision-Language Models (VLMs), high-resolution images produce a large number of visual tokens, resulting in high computational costs and KV-cache overhead during inference. To address this problem, we propose an Extreme Token Compression (ETC) framework that minimizes task loss when reducing the number of input tokens based on the principle of variational information distillation. Specifically, from an information-theoretic perspective, we show that minimizing task loss requires the compact representation to preserve the instruction-aware sufficient statistic of the task-relevant visual information for prediction. In practice, ETC leverages text-to-image cross-attention to weight the original visual features to approximate the latent instruction-aware predictive statistic. Moreover, ETC introduces a variational information distillation, enabling the compact representation to preserve the essential information to recover this predictive statistic. Experiments on LLaVA-1.5-7B and Qwen3-VL-2B show that ETC remains effective even under single-token compression, substantially reducing KV-cache overhead while retaining strong task performance.

Keep browsing

More papers from top AI conferences

30+ top conferences, by conference, year, and topic.