The paper introduces ESIA, an Earth Surface Immune System that detects and recognizes unknown anomalies in satellite imagery without prior category knowledge. It uses a non‑specific innate stage for rapid localization and a specific adaptive stage that matches image patches to text prompts via a multi‑modal model, achieving high F1 scores. The system adapts to new scenes in seconds and has been validated on a large global dataset, with applications to farmland degradation after the Kakhovka Dam collapse and burn severity assessment from the 2025 Palisades Fire.
By Jingtao Li, Qian Zhu, Xinyu Wang, Deren Li, Liangpei Zhang, Yanfei Zhong
The review surveys 4D millimeter‑wave radar perception algorithms for autonomous driving, covering signal processing, object detection, semantic segmentation, motion estimation, occupancy prediction, and dynamic scene reconstruction. It organizes the field by perception tasks, discusses radar fundamentals, data representations, and quality‑enhancement methods, and compares radar‑only learning, multimodal fusion, and cross‑modal supervision. The paper also summarizes datasets, annotations, evaluation protocols, and outlines common challenges and future research directions.
By Xumin Wu, Jun Zhou, Jilin Mei, Chen Min, Yu Hu
SceneTeract is a verification interface that separates semantic action understanding from physical feasibility in indoor 3D scenes. It decomposes activities into atomic actions and performs explicit geometric checks to determine executability, providing diagnostic traces for failures. The system reveals widespread functional and accessibility issues in synthetic scenes, shows that existing VLMs over‑predict action feasibility, and improves VLM performance through post‑training with verifier feedback, with benefits that generalize to real‑world scenes.
By L\'eopold Maillard, Francis Engelmann, Tom Durand, Boxiao Pan, Yang You, Leonidas Guibas, Maks Ovsjanikov
The paper presents a multimodal information system that integrates structured clinical summaries, medical images, and biomedical named entities from 52,949 open‑access case reports published between 2000 and 2021. This information is organized within a well‑structured medical ontology, and a user interface is designed to help junior clinicians search and browse case reports efficiently. The system aims to improve the identification and diagnosis of rare diseases by overcoming the limitations of traditional keyword‑based retrieval on unstructured case reports.
By Shuyu Guo, Lan Huang, Yichen Liu, Hanbin Ma, Tian Bai
The paper introduces TrioRAG, a graph-free multimodal retrieval-augmented generation framework that combines evidence from the question, an anchor image, and a VLM-enhanced query via late fusion. It also presents AutoQA, a benchmark featuring noisy web-sourced images that require reasoning across manuals. TrioRAG outperforms graph-based systems on three benchmarks while cutting costs and speeding up inference by 1.6–2.3×.
By Tithi Rakshit, Hongkuan Zhou, Lavdim Halilaj, Yuqicheng Zhu
The paper introduces Redemption Score (RS), a multi‑modal evaluation framework for image captioning that combines three complementary signals: Mutual Information Divergence for global image‑text alignment, DINO‑based perceptual similarity of cycle‑generated images for visual grounding, and LLM text embeddings for contextual similarity to human references. RS fuses these signals to provide a more holistic assessment, achieving a Kendall‑τ of 58.42 on Flickr8k and outperforming most prior methods. The framework demonstrates consistent performance across Conceptual Captions and MS COCO, offering a robust evaluation that captures both visual accuracy and text quality.
By Ashim Dahal, Ankit Ghimire, Saydul Akbar Murad, Nick Rahimi
Adversarial patches applied to Vision‑Language‑Action (VLA) policies not only corrupt actions immediately but also leave lasting state effects that persist after the patch is removed. The study introduces a state‑restoration protocol that evaluates recoverability after patch removal, distinguishing true adversarial impact from occlusion or action‑error magnitude. Experiments on OpenVLA‑OFT with EDPA attacks show that only 36.2% of episodes recover after five chunks, while controls recover at much higher rates; a recovery adapter can improve recovery but its effectiveness drops with delayed intervention.
By Enhao Wu, Fusen Guo, Yuxin Cao, Ziyang Lyu, Lin Li, Wei Song
BINDER is a new probabilistic model for medical image registration that builds on mutual information and uses latent voxel‑wise correspondences to enable closed‑form iterative updates. The approach yields a demons‑like optimization algorithm that performs robustly on both monomodal and multimodal tasks, and a sampler that quantifies uncertainty in high‑dimensional 3D deformations. The authors provide the code on GitHub for public use.
By Stefano Cerri, Amirhossein Hassankhani, Ya\"el Balbastre, Koen Van Leemput
CitySTAR introduces a training‑free framework that transforms billion‑scale urban point clouds into a query‑ready scene graph of open‑vocabulary 3D instances, using CodeLLM‑driven tools to supply multimodal evidence for node attributes and spatial relations. It models target‑context topology with paired hypergraphs and performs bidirectional topology verification for structural disambiguation, followed by a Reflective Cross‑modal Grounding module that integrates topology consistency and 2D visual evidence to decide over a metric‑aware 3D context graph. The authors also present CitySTAR‑3D, a benchmark that enhances semantic coverage, instance completeness, bounding‑box fidelity, and spatial‑relation complexity for city‑scale 3D grounding, and report extensive experiments showing consistent improvements in open‑world urban 3D grounding with strong interpretability and generalization.
By Shuai Zhang, Hongye Hou, Qinghe Liu, Zhuoxiao Li, Dongli Wu, Jing Ou, Yuan Liu, Wufan Zhao
The paper presents a new calibrated confidence framework for straight‑through processing (STP) of financial documents, decomposing confidence into perception, layout, and validation channels and applying conformal risk control. Using Vision‑Language Models (Qwen3.6‑27B and Gemini‑3.1‑Flash‑Lite) on three public datasets, the method raises AUROC from 0.54‑0.74 to 0.90‑0.99 and increases auto‑approval rates to 49‑72% while keeping empirical error below the target. This approach enables reliable, large‑scale STP without human review.
By Yichao Jin, Yushuo Wang, Yuxuan Han, Kwan Ching Yee Sonia, Weiyang Song, Chiu Jin-Chun Kent, Wong Chong Hwee, Wong Tiong Kiat, Kenneth Zhu Ke, Jingyuan Zhao
The paper introduces Oph‑Guid‑RAG, a multimodal retrieval‑augmented generation system tailored for ophthalmology clinical question answering. It treats each guideline page as an independent evidence unit, retrieving page images to preserve tables, flowcharts, and layout, and employs a controllable retrieval framework with routing and filtering to reduce noise. Evaluated on HealthBench, the system outperforms GPT‑5.2 and GPT‑5.4 on hard cases, achieving significant gains in overall score and accuracy, and ablation studies confirm the importance of reranking, routing, and retrieval design.
By Shuying Chen, Sen Cui, Zhong Cao
PosteriorBench is a new benchmark that evaluates how well generative inverse solvers recover full posterior distributions rather than just a single reconstruction. It tests four physics-based inverse problems—Darcy flow inversion, Poisson source recovery, carbon capture and storage, and light transport material inference—using high-fidelity reference posteriors generated by rejection sampling and MCMC. The benchmark employs five metrics (posterior-mean error, posterior-standard-deviation error, maximum mean discrepancy, sliced Wasserstein distance, and radially averaged power-spectrum error) to assess pointwise accuracy, uncertainty, distributional alignment, and global frequency fidelity, revealing significant distribution-matching gaps in current solvers and highlighting the importance of neural operators, guidance weights, and generation noise for posterior-variance calibration.
By Jiachen Yao, Zi-Siang Hsu, Xi Deng, Aditi Gupta, Xin Ju, Sally M Benson, Gege Wen, Anima Anandkumar
The paper introduces BTS-CAFE, a federated domain generalization framework for respiratory sound classification that addresses stethoscope-induced shortcuts. It combines causality-inspired device-style interventions, counterfactual metadata augmentation, and gradient alignment to reduce style–content entanglement and promote device-invariant decision boundaries. Experiments on ICBHI and SPRSound datasets show a 3.69‑point improvement in out-of-distribution performance over the baseline and outperform conventional data augmentation and federated learning methods.
By Heejoon Koo, Yoon Tae Kim, Miika Toikkanen, June-Woo Kim
The paper introduces GPUB, a large-scale benchmark for grounded product understanding in e‑commerce livestream videos, featuring 3,000 livestreams, 31K fashion products, and multi‑moment temporal annotations. It defines three evaluation tasks, with the main task (GPrU) requiring simultaneous product identification and moment localization. Existing multimodal models perform poorly on GPrU, prompting the authors to develop UniPro, which improves performance by learning product‑aligned, temporally structured representations.
By Xinyu Zhang, Junjie Chen, Jiawei Ge, Qianlong Li, Libin Ma, Baokun Pan, Yahui Luo
The paper introduces REVEAL, a diagnostic benchmark that stresses Video‑Language Models (VidLMs) on five controlled probes—camera‑motion sensitivity, cross‑frame integration, video sycophancy, language‑only shortcuts, and temporal expectation bias—to assess how well these models encode and use visual evidence. Experiments on 12 VidLMs reveal systematic failures: some visual signals are never reliably encoded, while others are overridden by model priors, leading to performance below chance on several probes that humans solve with high accuracy. Mechanistic probes further pinpoint where and why visual evidence is lost, demonstrating that under assertive prompts a model’s output becomes nearly invariant to real versus random video input, rendering visual evidence causally inert.
By Sethuraman T V, Savya Khosla, Aditi Tiwari, Vidya Ganesh, Rakshana Jayaprakash, Aditya Jain, Vignesh Srinivasakumar, Onkar Kishor Susladkar, Srinidhi Sunkara, Aditya Shanmugham, Rakesh Vaideeswaran, Abbaas Alif Mohamed Nishar, Simon Jenni, Rohan Maheshwari, Derek Hoiem
The paper introduces Biomedica, an open-source dataset sourced from PubMed Central that includes over 6 million scientific articles and 24 million image‑text pairs, along with 27 metadata fields and expert human annotations. To facilitate use, the authors provide scalable streaming and search APIs via a web server. They demonstrate the dataset’s value by training embedding models, chat‑style models, and retrieval‑augmented chat agents, all of which outperform previous open systems in their categories.
By Alejandro Lozano, Min Woo Sun, James Burgess, Jeffrey J. Nirschl, Christopher Polzak, Yuhui Zhang, Liangyu Chen, Jeffrey Gu, Ivan Lopez, Josiah Aklilu, Anita Rau, Austin Wolfgang Katzer, Collin Chiu, Orr Zohar, Xiaohan Wang, Alfred Seunghoon Song, Chiang Chia-Chun, Robert Tibshirani, Serena Yeung-Levy
The paper presents a method to automatically identify utterances in child-centered daylong audio recordings that can be reliably transcribed by modern ASR systems, enabling accurate transcription of a substantial portion of the speech. On four English corpora, the approach achieves a median WER of 0% and a mean WER of 16% when transcribing 30% of the total speech, compared to a median WER of 52% when transcribing all speech. Word frequency distributions from the automatic transcripts correlate strongly with manual annotations (r = 0.94 overall, r = 0.99 for frequent words).
By Daniil Kocharov, Azarias Galama, Okko R\"as\"anen
TacSushi is a tactile‑grounded, Cosmos3‑based world‑action policy for dexterous sushi manipulation. It encodes RGB, language, and hand state, fusing fingertip tactile data via feature‑wise gated fusion, and learns from future‑consequence predictions while excluding failed actions from imitation. Trained on 340 successful and 50 failed trials, TacSushi achieves 68.3% in‑distribution and 37.5% out‑of‑distribution success, outperforming baselines that lack future‑consequence supervision or use direct tactile concatenation.
By Haodi Hu, Kaen Kogashi, Toshiaki Koike-Akino
The paper introduces TAR (Traffic Anomaly Reasoning) and its evaluation suite TAR-Bench, designed to train and assess video‑language models on tasks beyond simple anomaly detection. TAR offers 44,040 chain‑of‑thought annotations covering 10 tasks across 3,670 CCTV videos, while TAR‑Bench supplies 960 human‑curated test annotations from 80 clips of 17 YouTube videos. Experiments show that strong question‑answering performance does not guarantee temporal or scene reasoning, and multi‑task fine‑tuning on TAR consistently improves model scores, with a 10‑task model outperforming its zero‑shot baseline by 21.4 points. The datasets serve as the official training and evaluation data for AI City Challenge 2026 Track 3.
By Han Zhang, Yilin Zhao, Zaid Pervaiz Bhat, Zheng Tang, Varun Praveen, Vidya N. Murali, David C. Anastasiu, Tomasz Kornuta
The paper introduces the workspace token, a lightweight latent memory representation for robotic manipulation that captures task-relevant historical information. By training with VLM queries only during training, the token can be queried efficiently at deployment, replacing full observations. Experiments in simulation and on hardware show that policies using the workspace token solve memory-intensive tasks without in‑loop VLM reasoning and even outperform heavier approaches.
By Nitish Dashora, Douglas Chen, Idan Shenfeld, John Marangola, Pulkit Agrawal, Max Simchowitz