Glass Surface Detection Grounded in 3D Visual Geometry proposes a new approach that grounds glass surface detection in 3D visual geometry rather than relying solely on 2D appearance cues. The method uses a visual geometry grounded transformer (VGGT) to distill 3D priors and creates glass-aware 3D representations, then applies a multi-task learning framework with a Frequency Self-Attention Module (FSAM) and a Geometry Grounding Block (GeGB) to localize and segment glass surfaces. Experiments show state‑of‑the‑art performance on seven benchmarks, good generalization to video and multi‑modal data, and significant improvements in reconstruction of glass scenes.
By Yiwei Lu, Ke Xu, Tao Yan, Xiaojun Chang, Radu Timofte, Rynson W. H. Lau
SOCO is a new benchmark for Semantic Object Correspondence that introduces a taxonomy of correspondence types and provides consistent, functionally meaningful keypoint annotations across 100 categories and over 1M correspondence pairs. It also includes keypoint language descriptions, enabling evaluation of large vision‑language models and their fine‑grained part‑level understanding. Experiments show that vision foundation backbones encode strong semantic structure but transfer correspondences poorly across related categories, LVLMs excel at text‑prompted part localization but lag in visual‑reference matching, and correspondence performance predicts dense downstream tasks more strongly than ImageNet classification.
By Olaf D\"unkel, Basavaraj Sunagad, Haoran Wang, David T. Hoffmann, Christian Theobalt, Adam Kortylewski
The paper introduces Text-to-Seed (T2S), a training‑free framework for open‑vocabulary semantic segmentation that repurposes Stable Diffusion to generate attention‑based seed points from text queries. These sparse seeds serve as point prompts for the Segment Anything Model (SAM), enabling reliable region expansion without relying on inaccurate coarse masks. T2S achieves strong performance on standard OVSS benchmarks using only the text‑to‑region correspondence of diffusion models and no task‑specific training or extra annotations.
By Kumju Jo, Heesun Jung, Sungyong Baik
HullWake is a new maritime vessel detection framework that prioritizes hull detection before wake analysis to address the wake-reliance problem. It separates hull evidence from wake context, extracts wake cues via bidirectional proposal-anchored corridors, and suppresses wake-dominant predictions through multiple supervisory strategies. The method is evaluated on a wake-oriented dataset and demonstrates improvements in overall AP, robustness to weak or no-wake vessels, reduction of wake-like false positives, and stability of confidence after wake attenuation.
By Yefan Wang, Xingyu Wang, Ruibiao Zhu, Yusen Wu
The paper introduces DPA-I2P, a depth-guided projective alignment method for image-to-point-cloud registration in autonomous driving. It employs Ray-Conditioned Metric Depth Encoding and Projection-Consistent Vision Lifting to align depth and visual cues geometrically, and uses Cross-Modal Query Pruning to enhance matching stability. Experiments on KITTI and nuScenes show significant reductions in rotation and translation errors compared to existing implicit baselines.
By Wenxin Zhang, Hang Li, Zhiwei Xu, Qiankun Dong, Gang Wang, Tao Li
The paper introduces G2D, a training‑free framework that combines a discriminative model (CLIP) for broad candidate retrieval with a generative vision‑language model for fine‑grained, image‑grounded verification. By using CLIP’s top‑K shortlist and a structured prior from candidate names and probabilities, G2D focuses generative reasoning on uncertain samples, achieving an average accuracy of 68.85% across eight benchmarks—higher than both CLIP alone (59.35%) and the standalone generative model (63.11%). The approach also adapts to various generator configurations and extends to other models such as DCLIP, WaffleCLIP, and CuPL.
By Zehua Hao, Fang Liu, Qinliang Wang, Yaoyang Du, Xinyan Huang, Puhua Chen
The paper introduces the Finance‑Aware Graph Spatio‑Temporal Network (FA‑GSTN) for forecasting realized volatility by treating the implied volatility surface as a dynamic graph. Nodes represent grid points on the surface, with edges capturing adaptive intra‑day spatial and explicit inter‑day temporal relationships, while finance‑aware node features (e.g., option Greeks) and a multi‑scale temporal smoothing gate address high‑frequency noise. Experiments on a large equity options dataset show FA‑GSTN achieves state‑of‑the‑art predictive accuracy (R² up to 0.473) and outperforms Vision Transformer baselines even with only one year of training data, demonstrating robustness during market stress.
By Chuanzhen Wang, Alice Zhang, Wei Chen, Michael Brown
The paper introduces an unsupervised domain adaptation framework that aligns redundancy-reducing features to enable accurate 3D segmentation of cone-beam CT (CBCT) without target-domain annotations or inference-time adaptation. The method is architecture-agnostic, working with both CNN-based and ViT-based foundation models, and is evaluated on two liver segmentation benchmarks for interventional vascular procedures and radiation therapy. Results show that even large pretrained segmentation networks need explicit feature-space bridging to generalize across diagnostic CT and CBCT, and the proposed approach consistently outperforms existing pretrained foundation models and UDA strategies.
By Gauthier Miralles, Loic Le Folgoc, Vincent Jugnon, Pietro Gori
The paper demonstrates that a self‑supervised Vision Transformer (ViT) pretrained on a fast, low‑cost semi‑numerical simulator can produce data summaries that transfer across different simulators without retraining. In 21cm cosmology, the ViT—named SKATR—pretrained on 67,000 21cmFAST lightcones is applied unchanged to hydrodynamical Loreli II lightcones, enabling accurate inference of five astrophysical parameters with fewer radiative‑transfer simulations than a fully‑supervised baseline. SKATR remains accurate, informative, and calibrated even under realistic SKA antenna array noise, outperforming supervised models retrained on noisy data.
By Yannic Pietschke, Caroline Heneka, Ayodele Ore, Romain Meriot
This systematic literature review examines 97 studies on optical character recognition (OCR) from 2015 to 2025, tracing the evolution of AI models, application domains, data types, and linguistic coverage. It identifies key OCR models, evaluates their performance, strengths, and limitations, and highlights unresolved challenges such as limited resources for underrepresented languages, high variability in handwritten text, and constraints in real‑time applications. The review proposes promising approaches—including self‑supervised learning, multimodal AI, AutoML, AI‑assisted postprocessing, TinyML, and joint corpora creation—to enhance OCR accuracy and address these challenges for industrial use.
By Nuzhat Khan, Ab Al-Hadi Ab Rahman, Shahriyar Masud Rizvi, Ibrahim Yousef Alshareef, Muhammad Nadzir Marsono, Muhammad Paend Bakht, Mohd Shahrizal Rusli, Shahidatul Sadiah
The study presents a new approach to detect Christmas tree plantations in high‑resolution aerial imagery, treating the task as a rare‑target semantic segmentation problem. It introduces a Hard Negative Mining strategy that significantly improves precision‑recall performance, achieving an IoU of 0.733 and an F1‑score of 0.846 on a 2020 test set. Temporal transfer experiments demonstrate the model’s ability to generalize across years, while large‑scale validation highlights the challenge posed by the plantations’ small spatial footprint.
By Francesca Razzano, Emanuele Dalsasso, Adrien Baysse-Lain\'e, Silvia Liberata Ullo, Gilda Schirinzi, Jocelyn Chanussot
The paper introduces UniFLM, a unified framework for segmenting and measuring fetal long bones in ultrasound images. It presents the Fetal Limb Bones (FLB) dataset with high‑quality annotations for the humerus, femur, tibia‑fibula, and radius‑ulna. UniFLM employs a Semantic‑Aware Skip Connection, a Positive Sampling strategy, and a Point Regression Mapping module to improve segmentation accuracy and bone length measurement, achieving superior performance over existing models on the FLB dataset.
By Zeen Zhou, Qiuhua Chen, Xiaojun Cao, Changmao Chen, Chao Sun, Bo Du
The paper introduces Successive Capacity Growth (SCG), a method for adaptively expanding Vision Transformer encoders in Joint-Embedding Predictive Architectures (JEPAs). SCG starts with a minimal encoder and incrementally increases width or depth based on a task‑agnostic test‑and‑verify mechanism, while a Sketched Isotropic Gaussian Regularizer (SIGReg) keeps learned semantic dimensions independent. Experiments on multi‑object dynamics and 2D navigation tasks show that SCG achieves up to 20.3% better prediction loss than fixed small baselines and 23% better than fixed large models, with far greater parameter efficiency and no false‑positive expansions.
By Frederik Berenz
The paper presents a method for cross‑architecture knowledge distillation from a fine‑tuned DINOv2 Vision Transformer teacher to a lightweight bidirectional Visual State Space Model (LVSSM) student for tea leaf disease classification. By addressing training‑stability issues with a progressive convolutional stem and gated selective‑scan block, the 4.45 M‑parameter student achieves a mean test accuracy of 95.41%—a 3.09‑point improvement over the teacher’s 92.32%—while using only one‑fifth of the teacher’s parameters. Ablation studies show that simple logit‑level distillation outperforms intermediate feature alignment, and the gains are specific to students that start below the teacher’s performance.
By Zibo Zhou, Zongsen Qiu, Rui Chen, Yujie Yao, Yue Zhou, Jianjun Wang
The study evaluates four deep‑learning segmentation architectures—Unet, PSPNet, Linknet, and FPN—paired with six pre‑trained encoders to predict COVID‑19 lesions in CT images. Experiments on three COVID‑19 CT datasets show high accuracy, achieving a maximum binary F1‑score of 98% and multi‑class F1‑scores of 75% and 77%. The work aims to provide a standardized performance benchmark for medical image segmentation and a reference for other imaging scenarios.
By Sarmad Khan, Basim Azam, Arslan Shaukat
The paper introduces TADP, a task‑aware deformable prediction framework for single‑stage 3D object detection. It employs a triple feature refinement aggregation module, a multi‑scale feature aggregation block, and a plug‑and‑play task‑aware deformation head to adaptively extract and fuse features for different detection tasks. Experiments on the KITTI dataset show that TADP achieves a car mAP of 80.91%, outperforming many state‑of‑the‑art methods.
By Su Wang, Yaochen Li, Min Yang, Jiaohao Nie, Chang Liu, Yuehu Liu
The paper introduces four new event‑based vision datasets created with the ANTShapes simulation tool, designed to support object classification research using spiking neural networks (SNNs). These datasets vary in difficulty and are benchmarked against established spiking datasets such as N‑MNIST, CIFAR10‑DVS, DVSGesture, and POKER‑DVS using a convolutional SNN. The work provides detailed, high‑quality datasets for future experiments and validates ANTShapes as a suitable tool for generating event‑based vision data.
By M. Middleton, H. Kayan, B. Sen Bhattacharya, T. Ali, E. Baikas, M. Vousden, C. Perera, O. Rhodes, E. Gheorghiu, M. A. Trefzer
FU-Mamba is a new framework for oralscan image segmentation that combines dynamic scanning with frequency domain enhancement. It introduces a Dynamic Mamba Block that learns adaptive sampling offsets for content‑aware scanning, preserving spatial coherence, and a frequency enhancement block that balances spectral components using wavelet‑guided decomposition and spectrum pooling. Experiments show a 1.1% improvement in mean intersection over union on a dental segmentation dataset.
By Xinxin Zhao, Jinpeng Ye, Bo Wei, Liqin Wu, Mahmoud Hassaballah, Karen Egiazarian, Aura Conci, Victor Hugo C. de Albuquerque, Abdulkadir Sengur, Leszek Rutkowski, Yan Tian
The paper argues that evaluating privacy‑enhancing technologies (PETs) solely through image classification is insufficient because classification remains robust to many geometric and local perturbations. It proposes a compute‑aware multi‑task protocol that uses lightweight proxy tasks to assess PETs across various transformations, revealing that PETs with similar classification accuracy can perform very differently on other vision tasks. The study demonstrates the necessity of broader evaluation metrics beyond classification to truly gauge PET effectiveness.
By Leon Ranke, Wolfgang H\"ubner, Ronny Hug, Michael Arens, J\"urgen Beyerer
PailitaoGR is a generative image retrieval model that incorporates a latent think-with-images approach to better handle real‑world query images. It uses a target‑focused perception mechanism—comprising a target enhancer and on‑policy distillation—to highlight the search target, and a selective auxiliary‑evidence mechanism—using an auxiliary enhancer and incremental contrastive distillation—to exploit useful side information. Trained on real‑world online image‑search logs, the method achieves an average 13.8 % improvement over existing baselines.
By Xiaomeng Fan, Yueran Liu, Shengyu Zhou, Chenghan Fu, Wanxian Guan, Feng Li, Chuan Yu, Jian Xu, Bo Zheng