VLM Evaluation

VLM: Vision-Language Model

Follow this topic Ask about this topic

Momentum

70 papers in the last four weeks, up 67% on the four weeks before. 0.7% of all new papers.

Jul 13Week of Sep 28

Latest papers 649

Sep 16, 2026cs.AI

MUSE: Benchmarking Large Vision-Language Models on Multi-Modal Understanding in Situated Education

Large vision-language models have achieved remarkable progress in multi-modal understanding, yet their capabilities in educational settings remain insufficiently evaluated. In AI-assisted language learning, models must interpret artistic imagery, understand its semantic, affective, and cultural content, and reason about visual context to support meaningful interaction. However, existing benchmarks primarily focus on real-world images or domain-specific educational reasoning, providing limited coverage of artistic educational content. To address this gap, we introduce MUSE, a benchmark for evaluating large vision-language models on artistic image understanding in situated educational applications. MUSE decouples image annotation from question generation, enabling diverse tasks with controllable difficulty while reducing annotation effort. It comprises twelve tasks spanning visual perception, semantic and affective interpretation, culture understanding, and compositional reasoning, together with diverse artistic images deliberately curated to center Singaporean and Southeast Asian multicultural contexts alongside Western art traditions, covering multiple themes and difficulty levels. Evaluation of open-source and proprietary models reveals substantial disparities across capability dimensions, particularly in affective interpretation and compositional reasoning. Our analysis further identifies common failure modes and key challenges for developing trustworthy multi-modal models for education. We hope MUSE will serve as a standardized benchmark for advancing multi-modal understanding in situated educational applications.

Luyao Zhu, Xun Wei Yee, Wei Li +2

AI Singapore, National University of Singapore, Singapore · Institute of Advanced Intelligence and Computing, A*STAR · School of Computing, National University of Singapore, Singapore

VLM Evaluation Large Vision-Language Models Cultural Knowledge in Language Models Multimodal Model Evaluation

Sep 16, 2026cs.CV

Decodable but Misrouted: Sparse Features Uncover a Readout Gap in Vision-Language Models for Harmful Meme Detection

When large vision-language models misclassify harmful memes, the failure may reflect missing internal evidence or an inability to route represented evidence to their outputs. We distinguish these cases in Gemma-3 and Qwen3.5 using sparse autoencoders, role-conditioned probes, causal interventions, and recovery experiments across six harmful content benchmarks, with additional Spanish and Hindi-English code-mixed evaluations. Sparse readouts outperform native prediction on all six primary binary tasks: Qwen averages

0.740

versus

0.432

for native macro-F1, residual reconstruction reaches

0.486

, and Gemma improves from

0.532

to

0.714

. These gains measure how accessible the label is to a supervised readout; they do not show that the model's native generation already applies such a decision rule. Under the evaluated scales, Qwen silent-feature ablation is

24-63

times more probe-sensitive, whereas routed-feature patching on literal yes/no tasks is

16-140

times more output-sensitive. Native-only threshold calibration explains much, but not all of the gap: on five tasks with matched probe scores, it recovers

69.8

% of the raw native-to-probe difference, while direct routing adds

0.094

mean macro-F1 beyond calibrated native scoring. Joint gold-label, probe-KL, and pairwise LoRA supervision improves dedicated FHM prediction, but a gold-only adapter performs better on the shared seven-task mean. A case study of Gemma-3-12B on the Facebook Hateful Memes dataset finds a distributed rank-32 image-prompt interaction, reaching

0.756

versus

0.685

native macro-F1. Robustness controls show that the signal is not explained solely by accompanying OCR and depends on paired visual evidence, and that it extends beyond English. In many of the errors we study, the evidence is represented but does not reach the answer; therefore, routing is a common bottleneck in harmful meme classification.

Girish A. Koushik, Diptesh Kanojia, Helen Treharne

Nature-Inspired Computing & Engineering Department of Computer Science University of Surrey Guildford, United Kingdom · Surrey Centre for Cyber Security Department of Computer Science University of Surrey Guildford, United Kingdom

VLM Evaluation VLM Robustness VLM Interpretability Classification

Sep 16, 2026cs.AI

The Mirage of Calibrated Confidence: Trajectory-Independence of Verbalized Confidence in Vision-Language Models

A calibrated Vision-Language Model (VLM) can repeatedly self-correct, say "Wait, I should recheck," arrive at the wrong answer, and still report high confidence. We find that this occurs because verbalized confidence is largely trajectory-independent in the VLMs and calibration methods we evaluate. We examine this through three complementary lenses: content variation, token masking, and the model's own hesitation markers. We show that confidence is insufficiently sensitive to what the reasoning trajectory actually contains, and that calibration training can paradoxically worsen this disconnect. Since existing metrics like ECE and AUROC cannot detect this problem, we propose the Trajectory-Grounding Score (TGS) in two complementary forms: TGS-self, which compares confidence with and without access to the model's own trajectory, and TGS-pair, which tests whether the model assigns higher confidence to correct trajectories than to flawed ones along the vision, reasoning, and answer axes. We propose TGS-Bench, a model-agnostic suite spanning 10 benchmarks with controlled good/bad trajectory pairs, and show that conventional calibration rankings diverge from trajectory-grounding rankings, exposing a blind spot in current evaluation practice.

Jisoo Yang, Jaeho Han, Trung X. Pham +1

Department of Artificial Intelligence, Chung-Ang University, Republic of Korea · Van Lang University, Ho Chi Minh City, Vietnam

VLM Evaluation Confidence Estimation in Language Models VLM Reasoning VLM Calibration

Sep 16, 2026cs.CV

EDCT-Bench: Uncovering Faithfulness Gaps in VLMs via Explanation-Driven Counterfactual Testing

Vision-Language Models (VLMs) can produce Natural Language Explanations (NLEs) that sound plausible yet remain inconsistent with the visual evidence they cite. We present Explanation-Driven Counterfactual Testing (EDCT), an intervention-based protocol that extracts visual concepts cited in a model's explanation, applies verified minimal edits to them, and tests whether the resulting answer and explanation remain consistent with the edited image. Using this protocol, we create EDCT-Bench, a comprehensive benchmark spanning three complementary domains: knowledge-intensive visual question answering (OK-VQA), safety-critical driving (DriveLM), and 3D spatial reasoning (3DSRBench). Across the evaluated VLMs, EDCT reveals substantial faithfulness gaps, with models frequently producing responses inconsistent with verified visual changes. Finally, our fine-tuning study suggests that EDCT-generated counterfactuals provide high-impact training signals.

Sihao Ding, Santosh Vasa, Aditi Ramadwar +1

Mercedes-Benz Research & Development North America

VLM Evaluation VLM Robustness Counterfactual Evaluation of VLMs Faithfulness of Language Model Explanations VLM Interpretability Faithfulness in VLMs

Sep 15, 2026cs.CV

Can VLMs Reliably Assess Sidewalk Accessibility Attributes from Pedestrian-Level Imagery?

An important component of urban accessibility, particularly for wheelchair users and people with reduced mobility, is sidewalk compliance with measurable requirements. We test whether effective width, longitudinal slope, cross slope, and pavement condition can be assessed reliably from pedestrian-level imagery using vision-language models (VLMs). We present the first application of sampling-based conformal prediction (CP) for VLM-based accessibility assessment. We evaluate four VLMs on 514 sidewalk images from Seoul, South Korea, with field-measured ground truth. Conformal calibration attains the nominal 90% coverage for all models and attributes, but the calibrated regions differ in informativeness. Effective width yields the most informative estimates, with a mean interval half-width of about 1.0 m for the best model. Since every model overestimates width, asymmetric calibration shortens the intervals by up to 33% at unchanged coverage. Longitudinal slope is marginally informative, cross-slope intervals are too wide to resolve regulatory thresholds, and pavement-condition sets degenerate to all five grades (A-E) for three of the four models. Uncalibrated intervals from raw sampling dispersion cover only 17-47% of field-measured values at a nominal 90% level. Among the images with the most self-consistent responses, these intervals miss the field-measured value in up to 96% of cases. Response self-consistency is therefore not evidence of accuracy, and sampling dispersion cannot be interpreted as uncertainty until it has been calibrated against field-measured ground truth. No quantitative attribute reaches the precision required for general compliance assessment, but CP identifies from calibration data alone which attributes can support screening of segments far from the thresholds. We release the annotated pedestrian-level images and their corresponding field-measured attribute values.

Seung Jae Lieu, Diego Morra, Chiara Cadoni +3

Senseable City Lab Massachusetts Institute of Technology Cambridge, MA 02139

VLM Evaluation Urban Planning VLM Calibration

Sep 15, 2026cs.CL

Vroom-Vroom at SHROOM-Visions: A Multi-Judge Committee for Detecting Hallucinated Spans in Vision-Language Outputs

This paper describes our submission to the SHROOM-Visions shared task on detecting and classifying hallucinated character spans in vision-language model outputs across four languages. We employ several fine-tuned vision-language models as independent annotators and combine their span predictions through character-level majority voting, and additionally explore activation probes. The approach ranks first in three of four languages and places on the podium in every language and metric. Our analysis indicates that disagreement among diverse models tracks disagreement among human annotators.

Toqeer Ehsan, Nico Penttilä, Richard Schmidt +2

Reliable Intelligence Team, Physical AI, VTT Technical Research Centre of Finland

VLM Evaluation Hallucination Detection in VLMs Hallucination Detection Multilingual VLM Evaluation

Sep 15, 2026cs.CV

Symmetry-Aware Likelihood-Orbit Aggregation for Selective Left-Right Claim Verification

Frozen vision-language models (VLMs) remain unreliable on fine-grained left-right claims, and raw claim likelihoods need not reliably rank verification errors. After a horizontal-reflection intervention is fixed, how should its induced likelihood measurements be combined into a selective verification signal? We introduce Relation-Orbit, a closed-form contrast with no learned fusion parameters that assigns eight normalized likelihoods to query-supporting and counterfactual roles determined by reflection, inverse relation, and entity exchange. A claim is asserted only when the signed contrast exceeds a threshold selected on held-out data using pointwise Clopper-Pearson upper confidence bounds. On VSR and GQA across four frozen VLMs, Relation-Orbit yields higher mean test coverage at a 10% selective-risk calibration target than an all-eight Orbit-Max baseline in all eight dataset-backbone settings; gains over a nearly abstain-all one-sided intervention score are reported separately. A separate LLaVA-1.5/COCO evaluation, reduced-orbit controls, and a two-sided partition diagnostic further characterize the structural advantage.

Zhouzhi Xiong, Chuxi Zhang, Weizhen He +3

Polytechnic Institute, Zhejiang University, Hangzhou, China · College of Artificial Intelligence, Zhejiang University, Hangzhou, China · College of Electrical Engineering, Zhejiang University, Hangzhou, China +1

VLM Evaluation Visual Spatial Reasoning Selective Prediction

Sep 15, 2026cs.CV

EgoPathBench: Evaluating Zero-Shot Egocentric Waypoint Decision-Making in Vision-Language Models

Zero-shot waypoint navigation requires vision-language models to select, from the current first-person observation, a sequence of spatial actions that is feasible for the agent and reaches the goal, placing joint demands on the integrated spatial intelligence of today's foundation VLMs. Existing spatial-intelligence benchmarks primarily evaluate isolated judgments of relations, directions, or targets and therefore do not directly measure the integrated navigation ability required to combine target recognition, action-consequence assessment, distance estimation, and path planning. To fill this evaluation gap, we introduce EgoPathBench, a dataset and five-task benchmark for first-person waypoint decision-making. Each question presents an egocentric RGB image, a natural-language goal, and numbered visible waypoints; a model returns traversable candidates or an ordered route. Predictions are evaluated for candidate feasibility, adjacent-edge legality, and goal arrival under point-agent or embodied geometry. EgoPathBench contains 31,852 training, 1,345 validation, and 1,111 benchmark questions and retains at least one geometrically verified reference route for every route question. Across nine VLMs, the highest EgoPath Score is only 28.3. The top-ranked model reaches 35.9% success on Point Path, but only 2.9% and 4.0% on Embodied Path and Intent Path, respectively, showing that current models remain limited in forming complete, goal-consistent routes under embodiment constraints. Beyond the evaluation data, we release the corresponding training resource. Fine-tuning Qwen 3.5 4B on the released training split raises its EgoPath Score from 3.9 to 38.9 and improves all four reported evaluations across three external spatial benchmarks, with gains of 1.4--9.6 points.

Yang Zhao, Zhuo Chen, Xubo Yang

Shanghai Jiao Tong University

VLM Evaluation Spatial Reasoning Benchmarks Robot Navigation Vision-Language Navigation

Sep 14, 2026cs.AI

Beyond Accuracy: Robustness, Cost, and Governance Trade-offs for Vision-Language Models in Templated Document Extraction

Vision-language models (VLMs) are increasingly used to extract structured fields from business documents, yet most evaluations report accuracy on clean benchmarks and offer little guidance to practitioners choosing an approach for a given task complexity. We address this gap with a measurement-grounded study and an open-source release. Across eleven systems (three commercial, two reasoning, five open-source VLMs in pretrained and fine-tuned form, and a non-LLM OCR->regex floor) scored on a 750-document held-out pool of synthetic checks, fine-tuning on 3K samples lifts the best open-source VLMs above F1 0.98-above every zero-shot commercial system on this task-while GPT-5 leads the commercial pool on F1 and Claude Sonnet 4.5 collapses on Date. To turn these measurements into actionable choices, we introduce a practitioner-oriented selection framework that maps a task profile (quality, latency, governance, volume) to a recommended approach via filtering and total-cost minimization, illustrated on a hypothetical mid-volume document-extraction scenario.

Kushal Patel, Pushkal Shrivastava, Mackenzie Lees +4

John Hancock, 200 Berkeley St, Boston, MA 02116, USA

VLM Evaluation VLM Robustness Document Information Extraction

Sep 14, 2026cs.AI

NoteVQA: Benchmarking VLMs on Real-Life Questions from Human Communities

Vision-language models (VLMs) increasingly power consumer-facing AI search, yet evaluating them on the diversity of everyday visual questions remains challenging. Existing benchmarks often target predefined capabilities, such as multi-hop retrieval or long-form synthesis, whereas users ask photo-grounded questions spanning a long tail of everyday scenarios. Despite advances in VLMs, users on Xiaohongshu, a mainstream Chinese image-sharing platform, continue to turn to other people for help with everyday visual questions. Motivated by this behaviour, we curate NoteVQA from these questions, yielding 252 items across 12 topical categories and 7 user intents. Each item includes a concise reference distilled from expert community responses and a human-audited interleaved reference answer that combines textual explanations with supporting visual evidence. We evaluate both short-answer correctness and interleaved-answer quality. To support the latter, we introduce AgenticInterleave, a single-agent ReAct framework for retrieval-supported answer generation, together with IVR-12, a 12-dimensional rubric for assessing the content, presentation, and image quality of interleaved references and model outputs. Across 10 frontier VLMs, the highest short-answer accuracy is 52.8%, while adding agentic search to Qwen3.5-397B-A17B improves accuracy by only 2.0%. For interleaved answers, the same model running AgenticInterleave scores 3.52 under IVR-12, compared with 4.65 for the human-audited references, with the largest gap in content quality. These results highlight the challenges that everyday visual questions pose for current VLMs in both answer accuracy and the quality of visually grounded explanations.

Haonan Jiang, Guojian Zhan, Jiancong Xie +6

AllSpark Team

VLM Evaluation Visual Question Answering Tool-Using Vision-Language Agents Multimodal Generation

Sep 14, 2026cs.CV

ModaLens: Measuring Image Sensitivity in Report-Conditioned Medical VLMs

A radiology report can already answer a clinical question, so it is hard to tell whether a vision-language model also uses the image. ModaLens, a paired image-swap audit, measures how report availability changes image sensitivity: MedGemma-27B on 3,199 paired MIMIC-CXR cases from 293 patients, all 14 questions per case (13 finding-specific and one composite), each image replaced by one from another study, usually of the same patient, with question and report fixed. Under an explicit answer instruction, the model's generated answer changes on 4.26 percent of trials with the report and 20.94 percent without it, a paired increase of 16.7 points (patient-clustered 95 percent CI 15.6 to 17.7), so report availability reduces image-swap sensitivity under this protocol; the original prompt with a lowercase first-token readout gives 4.70 percent against 17.07 percent, and substitutions also move continuous answer scores where the binary prediction does not change. The labels are derived from reports, which limits conclusions about visual correctness; the direction replicates in two further model lineages. Code, the exact prompts and a run record for every number are at https://github.com/criticaldata/MODALENS.

Sebastián Andrés Cajas Ordóñez, Maximin Lange, Quang Bui +9

MIT Critical Data, Massachusetts Institute of Technology · King’s College London · American International School Vienna +6

VLM Evaluation Radiology VLMs Sensitivity Analysis

Sep 13, 2026cs.CV

Vision-Language Models for Criterion-Level Grading of Handwritten Examinations in Outcome-Based Education

Criterion-level grading connects examination performance to learning outcomes, but manual marking introduces workload and variation between markers. This study evaluates vision-language models (VLMs) for handwritten outcome-based assessment across five dimensions: accuracy, human agreement, repeated-run reliability, error concentration, and explanation quality. Using 1,982 criterion-level records from 485 undergraduate examination answers, we compare 20 configurations spanning Qwen2.5-VL, InternVL3, Pixtral, a Donut baseline, and a cascade ensemble. Evaluation setups include zero-shot prompting, few-shot prompting, partial fine-tuning, and Low-Rank Adaptation (LoRA). Two independent faculty markers regraded all 291 test criteria, providing a human agreement baseline on the same assessment materials. Qwen2.5-VL with LoRA achieved Quadratic Weighted Kappa (QWK) of 0.727 and mean absolute error of 0.435 marks against the examiner, compared with mean human-pair QWK of 0.551. This comparison reflects calibration to the examiner's training marks. LoRA outperformed partial fine-tuning for all three instruction-tuned VLMs, while few-shot prompting reduced QWK in every configuration with valid prompted scores. Aggregate reliability and exact repeatability diverged: intraclass correlations ranged from 0.790 to 0.874, yet 50.2-63.6% of criteria changed marks across five sampled runs. Attention-guided deletion showed no statistically significant advantage over random masking, and four faculty reviewers reached no consensus on explanation usefulness. These findings highlight the need for rubric-specific calibration, repeatable scoring, review of consequential errors, and separate validation of explanations. The released evaluation protocol supports criterion-level assessment research and grading tools with teacher oversight.

Asif Hasan Tonmoy, Saad Ahmed, Md Khalid Syfullah +1

VLM Evaluation Rubric-Based Evaluation VLM Calibration Automated Grading

Sep 12, 2026cs.CV

New Evidence, Same Choice: Testing Physical Experiment Selection in Vision Language Models

A model first sees an image from one physical measurement experiment, such as how far a block coasted, and must answer a question about a new trial, such as whether the block will pass a target after a fixed push. The initial experiment may provide enough information to answer, or the model may need another measurement, such as the object's mass, friction, restitution, or spring stiffness. We study whether vision language models can decide when to answer immediately and, when more evidence is needed, which experiment to perform. Current physical reasoning benchmarks usually evaluate only the final answer, so they do not directly measure this decision-making ability. We introduce a controlled evaluation where each problem provides one measurement image and four possible physical worlds created by combining two possible masses and two possible values of another relevant property. The model must either stop and answer or select the cheapest additional experiment that can resolve the question. We construct matched problem pairs where changing either the observed measurement or the question changes the optimal action. Since all possible worlds and experiment costs are known, we can explicitly determine the optimal choice. Across six open models and 144 physical parameter sets, direct responses repeat the same action for 95.1% to 100% of image pairs even when the correct action changes. Brief reasoning improves action switching, but the best model makes both decisions correctly for only 5.9% of image pairs. Additional analysis reveals failures in measurement interpretation, physical reasoning, and response formatting. By evaluating evidence selection separately from final answers, our benchmark reveals limitations in physical reasoning that conventional answer accuracy can overlook.

Sourajit Saha, Shubhashis Roy Dipta, Nobin Sarwar +4

University of Maryland, Baltimore County · University of Georgia · Independent Researcher

VLM Evaluation VLM Reasoning Visually Grounded Reasoning

Sep 12, 2026cs.CV

HALDETECT at ImageEval 2026 Shared Tasks: Answer-First Contrastive Grounding with QLoRA

Large multimodal models tend to hallucinate visual detail fluently, which limits their deployment for fine-grained interpretation. We present HALDETECT, our system for the English hallucination-detection track (Task 1b) of ImageEval 2026, in which a system must identify, from an image and three culturally plausible statements, the single visually grounded one. We frame the item as one contrastive decision, emit the answer before its explanation, and structure reasoning around colour/texture, shape/form, and context. Our best submitted adapter fine-tunes Qwen2.5-VL-7B-Instruct with 4-bit QLoRA while freezing the vision encoder and reaches Contrastive Instability (CI) 0.035 on the 1,000-item test set; we placed third of eight teams. Development experiments show that answer order can matter more than model scale and that adaptation beats prompting alone. Retrospective paired analysis of the released gold labels confirms the QLoRA gain over the best prompt but not the small gap between the devtest-selected and best-test adapters, and reseeding all four training sizes shows that the apparent data-scaling curve does not survive a seed change. The 35 residual errors are culturally plausible function, material, and recognition distinctions; naive adapter voting does not help.

Syed Mohaiminul Hoque, Md Sakhawat Hossain

Independent Author

VLM Evaluation Hallucination Detection in VLMs Vision-Language Models Audio-Language Model Hallucination Detection VLM Hallucination Vision-Language Grounding

Sep 12, 2026cs.AI

Can Edge-Deployable Vision-Language Models Identify Species?

Camera traps often run in the field on edge hardware with limited or no connectivity, making small, locally-deployable vision-language models (VLMs) -- not frontier-scale ones -- the practically relevant class to evaluate for species identification. We test whether models in this deployment-relevant 2--8B range carry genuine taxonomic knowledge, evaluating four such VLMs (Qwen3-VL 2B/4B/8B, Gemma3 4B) against the domain-specific specialist BioCLIP (300M parameters) on a 96-species task, comparing clean iNaturalist photographs against camera-trap imagery from 6 LILA.science collections, on two independently-sampled evaluation sets. All models identify species far above chance, but every model -- general-purpose or specialist -- degrades sharply on field imagery (domain gaps of 9.6--26.6 percentage points, consistent across taxonomic levels and both evaluation sets), indicating the degradation reflects general image legibility rather than fine-grained discrimination failure. BioCLIP substantially outperforms every VLM tested (by 33.2--59.2 percentage points across an expanded 200-image sample for every model) despite its far smaller size, suggesting the gap reflects specialized training data rather than model scale; yet BioCLIP's own domain gap (18.0 points) is statistically indistinguishable from the best VLM's (22.3 points), suggesting the clean-to-field degradation itself is a property of the image-quality shift rather than a general-purpose-model weakness. Under open-set prompting, 5.9--9.6% of responses are syntactically valid but taxonomically nonexistent species names; the relative fabrication-rate ranking across models replicates exactly across both evaluation sets, a more robust finding than any single point estimate.

William Zhou, Mayukha Siripuram, Xiao Yan +2

Plano West Senior High School

VLM Evaluation VLM Robustness Vision-Language Models Image Classification

Sep 11, 2026cs.CV

Evaluation of Vision-Language Models Across Diverse Coastal Environments

Vision-language models (VLMs) enable robotic per- ception by associating visual observations with natural-language concepts. Yet their performance in coastal environments remains largely unexplored. We introduce a densely labeled coastal dataset containing more than 1,000 images collected across seven missions in three regions of Oahu, Hawaii, with 18 semantic classes and over 7,400 annotated instances. We evaluate seven modern VLMs through three complementary experiments mea- suring text-to-mask, mask-to-mask, and mask-to-text alignment. Broad landscape classes are generally recognized more accurately than conventional object and coastal classes, with coastal con- cepts presenting the greatest challenge. However, comparisons of shared conventional classes across coastal and terrestrial datasets reveal no consistent performance difference attributable solely to environmental context. Mask-to-mask matching also remains similar across conventional and coastal classes, while alternative textual labels substantially improve recognition of several coastal concepts. These results suggest that lower performance on coastal classes (at least on the objects/query categories evaluated) is heavily influenced by segmentation and linguistic representation.

Seth Knoop, Chad R. Samuelson, Gabriel R. Slade +2

VLM Evaluation Open-Vocabulary Semantic Segmentation

Sep 10, 2026cs.CV

VANTAGE-Bench: Evaluating the Infrastructure AI Gap in Vision-Language Models

As Vision-Language Models (VLMs) advance toward physical deployment, the focus has remained on action-oriented Embodied AI evaluated on subject-centric consumer video. This overlooks a pervasive class of Physical AI: Infrastructure AI, which relies on fixed cameras for open-loop insights like safety monitoring and operational logging. We introduce VANTAGE-Bench, a benchmark measuring this "Infrastructure AI Gap." It spans three operational domains (Logistics, Transportation, and Smart Spaces), unifies image and video evaluation across semantic, spatial, temporal, and spatio-temporal capabilities, and moves beyond multiple-choice to eight task formulations including dense captioning and spatio-temporal grounding. It adds a single-pass trajectory protocol for Single Object Tracking and, to our knowledge, the first such evaluation on fixed-camera infrastructure video, scored against specialist trackers. Annotation spans three regimes over 3,346 media assets: 3,342 video-task annotations, 4,281 image-grounding annotations, and 27,404 detection boxes. Evaluating 17 models zero-shot, we find the shortfall relative to consumer-centric benchmarks is concentrated, not general. Event verification, referring expressions, and temporal localization fall roughly 9 to 24 points at every model scale, while video question answering stays within 5.3 points of VideoMME and 2D spatial pointing shows no shortfall against BLINK. The temporal pillar is weakest in absolute terms: no system exceeds 55.7 mIoU on temporal localization or 37.3 SODA_c on dense video captioning. On tracking, frontier models come within roughly 5 points of specialist trackers over short horizons but separate as the horizon extends. Open-weight models lead 2D object localization outright, so neither scale nor proprietary access explains the pattern. Data, evaluation harness, and leaderboard: https://vantage-bench.org/

Zaid Pervaiz Bhat, Nimra Nayyar, Arihant Jain +6

NVIDIA · Clemson University

VLM Evaluation Vision-Language Models Video Object Detection

Sep 9, 2026cs.CV

Spot-the-shift: Evaluating Grounded Image Difference Captioning of Long-term Changes

Long-term change understanding from images of the same place revisited over time is a challenging task with applications in map maintenance and urban infrastructure monitoring. Prior work addresses it either through pixel-level prediction or difference captioning, neither of which is sufficient to reliably measure how well models detect and describe such changes. We introduce SPOT-THE-SHIFT, a human-verified benchmark for grounded image difference captioning of long-term changes in real-world driving scenes. Our benchmark provides natural language captions and spatial masks for structural changes across each image pair. We further propose an evaluation protocol that reliably assesses models' captioning ability, validated through human studies. Benchmarking state-of-the-art MLLMs, we find that models struggle with the fine-grained multi-image spatial capability required for this task. Finally, we develop a synthetic data generation pipeline that improves an off-the-shelf MLLM without sacrificing general capabilities.

Benedetta Liberatori, Nermin Samet, Paolo Rota +4

University of Trento · Sorbonne Université · Fondazione Bruno Kessler

VLM Evaluation Autonomous Driving Benchmarks

Sep 9, 2026cs.CV

LogiScope-VQA: Benchmarking Vision-Language Models for Logistics Hazard Identification in Industrial Scenarios

Large Multimodal Models (LMMs) large-scale deployment in industrial warehouse settings specifically necessitates that models exhibit human-expert-level hazard-oriented perception, understanding, and reasoning capabilities. However, the scarcity of real industrial data, tightly coupled to commercial terms, significantly hampers further advancement. To bridge this gap, we curate LogiScope-VQA to investigate the practical applicability of mainstream LMMs in real-world logistics operations. LogiScope-VQA comprises 2,476 images and 2,918 videos primarily sourced from real-world logistics parks, along with 10,274 VQAs meticulously curated and validated by human annotators. Grounded in 18 core objects and 20 risk types, we devise 39 subtasks aligned with three principal themes: industrial element perception, warehouse knowledge understanding, and potential risk reasoning. Furthermore, we incorporate dynamic thinking-budget configurations and dual-dimensional risk bias analyses to elucidate the properties of LMMs. Extensive experiments unveil that even powerful proprietary models, including GPT-5.5, Gemini-3.1-Pro, and Claude-Opus-4.7, exhibit a significant gap relative to human performance. The unique challenge of jointly integrating perception, understanding, and reasoning for hazard identification poses substantial headroom for further improvement on LogiScope-VQA. We additionally reveal the pervasive security bias issue that impedes LLMs' practical deployment in real-world settings. The industrial dataset is publicly available under the CC BY-NC-SA 4.0 license.

Hanjing Zhou, Mingze Yin, Ying Lian +3

Cainiao Group, Alibaba Group · Zhejiang University

VLM Evaluation Visual Question Answering AI Safety

Sep 8, 2026cs.CV

Charts Are Beyond Pixels: Probing for Layer-Wise Chart Understanding and Editing

Charts are structured visual compositions whose elements have distinct functional roles, semantic correspondences, and visibility relations. This structural view motivates evaluating whether models can understand and manipulate charts at the layer level. Existing chart benchmarks, however, primarily assess the correctness or fidelity of final outputs and do not directly evaluate these layer-wise behaviors. We present LayerWiseBench, a benchmark organized around three core concepts, layer attribution, layer binding, and visibility ordering, that structure its chart-understanding and chart-editing evaluations. Generated from executable chart programs, LayerWiseBench pairs each rendered chart with spatially aligned per-layer RGBA assets and construction-derived labels for functional roles, semantic bindings, and visibility relations. From this layer-wise representation, we derive controlled understanding questions, editing targets, reference images, and evaluation regions. It contains 2,800 source charts across 14 chart paradigms, from which we derive 7,329 layer-wise understanding questions and 53,791 instruction-guided editing variants. Among the evaluated VLMs, Qwen3.5-27B, which achieves the highest QA macro-average, obtains 93.04% accuracy on layer attribution and 97.46% on layer binding, but only 61.46% on visibility ordering. Across the four evaluated image editors, overall mIoU ranges from 1.49% to 4.93%, and visibility-constrained edits have the lowest mIoU for every editor, ranging from 0.37% to 2.00%. Taken together, these results identify tasks involving front-to-back relations between overlapping components as a recurring challenge across understanding and editing, motivating more explicit modeling of component identity and visibility relations.

Xiaochuan Zhong, Yifan Hou, Chenxi Pang +1

DeepDelta Lab, School of Artificial Intelligence, Shanghai Jiao Tong University · ETH Zurich · Google DeepMind

VLM Evaluation Chart QA Chart Understanding Image Editing

Sep 8, 2026cs.CV

Human-Centric Image Captioning with Subject-Centered Spatial Understanding

While multimodal large language models (MLLMs) achieve remarkable performance on generic image captioning, they frequently suffer from structural hallucinations in human-centric scenarios. Accurately modeling human subjects is foundational for critical downstream applications, such as accurate avatar/video/image generation and fine-grained human action understanding. However, these tasks require highly precise subject-centered spatial grounding, such as distinguishing egocentric left/right laterality and maintaining correct anatomical-object bindings. Although catastrophic for structural integrity, these localized spatial inversions are often overshadowed by overall descriptive metrics in existing benchmarks. To systematically expose and quantify this bottleneck, we introduce SPACE (Subject-centric Poses, Appearance, and Characteristics Evaluation), a benchmark designed to evaluate subject-centered spatial understanding. On SPACE, we reveal that despite strong generic perception, current MLLMs consistently fail to ground descriptions in the subject's intrinsic frame of reference. To bridge this gap, we propose a specialized data construction and alignment pipeline. We first extract structured spatial hints from fine-grained body-part localization to guide a two-stage caption rewriting process, yielding highly spatially-faithful training data. Furthermore, we design a rubric-based reward for Group Relative Policy Optimization (GRPO) that explicitly penalizes structurally critical spatial errors during alignment. Extensive experiments on SPACE demonstrate our framework significantly improves human-centric caption quality, particularly in subject-centered spatial reasoning, achieving performance competitive with strong closed-source models. Our benchmark and code are available at https://github.com/JHang2020/SPACE-Eval.

Bozhou Li, Jiahang Zhang, Yue Ding +11

Peking University · Kling Team · Beijing, China +5

VLM Evaluation Image Captioning Vision-Language Grounding

Sep 8, 2026cs.CV

VI-Bench: Benchmarking Prompt Inversion from AIGC Videos

Recent advances in video generation have made prompt-based control increasingly central to AIGC video generation. Prompts specify what a video should depict and how it should be represented, controlling factors such as visual style or camera behavior. Understanding this recoverability is important both for creative reuse and editing, and for assessing prompt leakage risks. However, existing video understanding benchmarks do not measure this capability: a caption may describe what is visible, but a replayable prompt must recover the generation-relevant controls needed to reproduce the video. To address this gap, we introduce VI-Bench, a benchmark built from 16.1 million real-user prompts and 900 human-verified AIGC videos. VI-Bench spans three progressively harder settings, namely single-shot semantic grounding, control over style and camera behavior, and multi-shot compositional inversion, and evaluates five generation-critical dimensions: subject, action, scene, style, and camera. We evaluate 18 representative VLMs, including 2 proprietary and 16 open-source models on VI-Bench, using an Inversion Score that measures prompt-level alignment with the original prompt and video-level fidelity of the regenerated video. The results reveal substantial limitations: even the strongest model achieves only 0.632 on Inversion Score, performance degrades sharply as samples require richer control and multi-shot reasoning, and models often produce plausible prompts whose regenerated videos deviate from the reference. These findings show that video prompt inversion is a distinct and under-evaluated capability requiring models to transform visual understanding into replay-stable generative control.

Wulin Xie, Rui Zhao, Kecen Li +5

Institute of Automation, Chinese Academy of Sciences · University of Chinese Academy of Sciences · University of Virginia +3

VLM Evaluation Video Generation Evaluation

Sep 7, 2026cs.LG

I Don't Miss You, but I Do: Self-Explanation Faithfulness of Modality Missingness in Vision-Language Models

Vision-language models (VLMs) are increasingly used in settings where some input modalities may be unavailable, yet we know little about whether they can faithfully explain how such missing information affects their own predictions. We introduce an interventional protocol for evaluating self-explanations of modality dynamics: models state what each modality alone would support, whether restoring a missing modality would change their answer, and whether the available evidence is sufficient; we then execute the corresponding intervention and compare these claims with realized behavior. We evaluate ten VLMs spanning open-weight and proprietary models across four tasks covering mixed, redundant, and unique modality regimes. We find a systematic tendency to overstate the sufficiency of available evidence. Models substantially underestimate the effect of restoring missing modalities: executed change exceeds predicted change in 78 of 80 model-task-condition settings, with task-level median executed change rates reaching 70.1% while median predicted rates remain at most 9.6%. Insufficiency claims have low recall, leaving many cases in which behavior changes despite a stated claim of sufficiency. Retrospective self-explanations show the same tendency, over-crediting single-input sufficiency in mixed regimes and interchangeability in redundant ones. Together, these results show that VLMs systematically mischaracterize how their predictions depend on available and missing evidence, motivating executable interventions as a behavioral test of multimodal self-explanations.

Aydin Javadov, Daniel Schoess, Florian von Wangenheim

ETH Zurich

VLM Evaluation VLM Interpretability Faithfulness in VLMs

Sep 3, 2026cs.CV

Principia: Relational Physics Tests for Video Models

Evaluating physical reasoning in video models is difficult because absolute motion measurements depend on frame rate, object scale, and camera calibration, all of which are often ambiguous or unavailable in generated video. We propose a different approach. When two objects in the same scene obey the same physical law, their motions must satisfy predictable relationships, and these relationships hold independent of calibration. We introduce Principia, a benchmark that evaluates Newtonian physics through relational consistency between paired objects. Principia spans eight phenomena - gravity, restitution, friction, rotational inertia, projectile motion, momentum, pendulum, and mass-spring oscillation - across translational, rotational, collisional, and oscillatory dynamics, using real-world scenes recorded under controlled protocols. We also introduce a calibration-independent consistency score that quantifies physical violation directly in image space. Across thousands of generations from six state-of-the-art video generators, no model exceeds 0.42 on Principia despite all scoring around 0.8 on VBench. Vision-language models are evaluated on their ability to detect relational physics violations, with the best model achieving only 67% accuracy and most performing near chance level.

Varun Varma Thozhiyoor, Shivam Tripathi, Venkatesh Babu Radhakrishnan +1

Indian Institute of Science · Johns Hopkins University

VLM Evaluation Physical Consistency in Video Generation Physical Reasoning in Video Video Generation Evaluation

Sep 3, 2026cs.AI

SVG-Score: Human-Aligned Evaluation of Text-to-SVG Generation

Scalable Vector Graphics (SVG) generation is attracting increasing attention as generative models improve in expressiveness and controllability. Progress, however, is held back by the lack of domain-specific evaluation protocols: current practice relies on metrics designed for natural images, most notably CLIPScore, which was never trained on vector graphics and aligns only partially with human judgment. We introduce \textbf{\ours}, a human-aligned evaluation framework for text-to-SVG generation. Through controlled caption and image perturbations, we first show that CLIP-based scores barely react to the errors SVG generators actually make, such as wrong colors, counts, and spatial relations, and that off-the-shelf Vision-Language Model (VLM) judges, while more sensitive, respond unevenly across error types and SVG styles. We then introduce a human-annotated dataset for \textit{Semantic Alignment}, measuring how faithfully a generated SVG reflects its caption. Building on it, we develop two complementary evaluators: CLIP scorers adapted to vector graphics and then aligned to human preferences, for fast large-scale evaluation, and a VLM judge trained with supervised fine-tuning and reward-shaped reinforcement learning, for more expressive and interpretable assessment. Using both, we benchmark major open-source, commercial, and optimization-based SVG generators on an independent caption set.

Marco Cipriano, Leonardo Zini, Alexandra Schild +5

Hasso-Plattner Institute, Germany · University of Modena and Reggio Emilia, Italy

VLM Evaluation Human Preference Modeling T2I Generation

Sep 3, 2026cs.RO

FailBench: How Reliable are VLMs at Judging Robot Task Success?

Vision-Language Models (VLMs) are increasingly used to evaluate robot manipulation outcomes, but existing benchmarks offer limited evidence of cross-domain generalization. We introduce FailBench, a benchmark for robot failure detection comprising 2,197 manipulation attempts across 14 public sources (12 real-world, 2 simulated). In FailBench, 75% of failures occur naturally, and six real-world sources come from non-failure-detection datasets. Evaluating 13 VLM-based detectors, we find the best model achieves only 0.77 mean balanced accuracy. Notably, models fine-tuned for failure detection consistently underperform general-purpose VLMs and their own pretrained baselines. Performance depends heavily on required visual evidence: models approach saturation when outcomes depend on observable object motion, but degrade to near-chance (<0.60 balanced accuracy) on contact-intensive assembly tasks. Error analysis reveals a systematic bias toward predicting success under ambiguous evidence, which persists even with increased reasoning effort. Finally, we show that input-level intervention--spatially localizing and cropping outcome-relevant regions--improves the top detector by 2.4 percentage points without extra training.

Zaruhi Navasardyan, Tatul Danielyan, Hrant Davtyan

Metric AI Lab

VLM Evaluation Robot Failure Detection Robot Manipulation Benchmarks VLMs for Robotics Anomaly Detection

Sep 3, 2026cs.CV

When Do Frozen VLMs Respond to Image-Free Object-Token Edits? An Answer-Key-Free Protocol and What It Reveals

Answering what-if queries about a scene with a VLM usually means injecting the assumption as text or repainting the scene with a generative model. We instead move the edit to the representation level, before the model input. The image is abstracted into a set of object-level tokens, and the original image never enters the VLM. This design rests on an open question: when do frozen VLMs actually respond to such token edits? We introduce an answer-key-free protocol: no post-edit answer is annotated. It scores edits whose answers are logically determined, and audits itself by reversing each scoreable choice. The protocol reveals three structures. The response is not free: explicit edit teaching, not ordinary VQA training, produces it in dense scenes and multiplies it in sparse ones, on all three operations. Once on, it is governed by token cleanliness and density, with deployable detector+segmenter tokens competitive with the oracle and outperforming it on VRSBench. And reading is a separable axis: the image-free token route preserves 92-96% of a matched patch-token baseline's free-text VQA, and the answers measurably depend on the tokens. The response, cleanliness, and reading structures are sign-preserved across two remote-sensing datasets (iSAID, VRSBench) and three frozen LM backbones. We release the probe generator, records, judge logs, and code.

Wonbin Son, Gyumun Choi, Junil Seo +3

Changwon National University · Chung-Ang University

VLM Evaluation Counterfactual Evaluation of VLMs VLM Reasoning

Sep 3, 2026cs.CL

FPCO-Dialog: A Multi-Turn False-Premise Benchmark for Correction and Cooperation in Vision-Language Models

Vision-language models (VLMs) are increasingly deployed in multi-turn settings where users may describe visual content with incorrect assumptions. Yet existing evaluations rarely isolate how models respond when the same visually grounded false premise persists across dialogue turns. We introduce FPCO-Dialog, a benchmark for evaluating correction and cooperation behavior in VLMs under repeated false premises. FPCO-Dialog contains 1,080 images and 10,800 question turns, stratified by visual complexity, object category, and false-premise class, and uses a 10-turn protocol in which a correct dialogue prefix is followed by repeated false-premise referring expressions. We evaluate 20 commercial and open-source VLMs with a model-agnostic protocol and CorrTP@K, a correction-rate metric over false-premise turns, scored by two independent detectors. FPCO-Dialog reveals substantial and persistent cross-model differences in aggregate correction tendency, model-specific turn-wise dynamics, and systematic variation across false-premise types under the benchmark's substitution distribution. The dataset, evaluation protocol, model outputs, detector labels, and code are available.

Jiayuan Ma, Yuqi Lu, Weiyang Guo +5

Harbin Institute of Technology, Shenzhen, China

Sep 2, 2026cs.CV

Seeing the City or Recognizing the Place? What Street-View Imagery Adds Beyond Existing Urban Data in VLM-Based Urban Sensing

Street-view imagery is increasingly analysed with vision-language models (VLMs) to infer urban attributes, but predictive accuracy alone does not show how much a photograph contributes beyond data already available for the same place. Using three VLMs, we compare image-based predictions with existing urban data for seven attributes drawn from five public resources. Each urban unit is evaluated with imagery, with location or text context, and against non-image predictions from nearby observations or public records. We also replace images and add conflicting records to test which source the predictions follow. Nearby observations or public records matched or exceeded image-only predictions for road damage, curb ramps, population, and house price. Images were more informative for building type, building function, and the floor count of low-rise buildings. For floor count, the advantage of images over nearby OpenStreetMap labels increased by 5.7 percentage points per doubling of distance to the nearest labelled building and declined for tall buildings whose rooflines often fell outside the frame. When images and records disagreed, predictions usually moved towards the supplied record. Released OpenFACADES floor annotations, generated with OpenStreetMap floor values as input, showed the same dependence: their agreement with the reference increased with building height, whereas that of image-only reruns decreased. The value of street-view imagery therefore depends on whether an attribute is visible and how well the place is already covered by existing data. Because machine-derived labels are often reused as references, these comparisons also bear on how urban datasets are documented and evaluated.

Kaizhen Tan

New York University

Sep 2, 2026cs.CV

Who Drives the Probability Game of VLMs? A Temporal Causal Drive Evaluation Framework

Vision-language models (VLMs) are increasingly evaluated on complex image and video understanding tasks, yet conventional metrics primarily assess final-answer quality and reveal little about how different information sources shape the generation process. We propose a causal and temporal evaluation framework that traces the evolving roles of visual input, question text, and generated prefixes during autoregressive decoding. Grounded in a Structural Causal Model, we use interventions and backdoor adjustment to derive three step-indexed causal-drive metrics---Visual Causal Drive (VCD), Question Causal Drive (QCD), and Prefix Causal Drive (PCD)---for characterizing source-specific generation patterns without requiring reference answers. Experiments on Qwen3-VL-8B-Instruct across MAVIS, LLaVA-Video-178K, and MiraData, together with cross-model validation on InternVL2-8B, reveal a consistent transition from stronger early question and visual guidance toward increasing reliance on generated prefixes. Randomized-intervention validation shows that QCD and PCD reduce recovery error over observational PMI baselines by 34.8% and 47.1%, respectively. On VLMBias, the prefix--visual imbalance score achieves 0.767 AUROC and 0.873 AUPRC for distinguishing prior-driven from visually grounded generations. These results show that causal-drive trajectories provide complementary source-level diagnostics for multimodal generation.

Shuyao Xiao, Shengling Wang, Haoyu Niu +4

College of Artificial Intelligence, Beijing Normal University · Kuaishou Technology

VLM Evaluation VLM Interpretability Causal Interventions in Language Models