Peer Revenue Fully funded

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

Computer Science Computer Vision

cs.CV Watch Computer Vision

Monday 13 April

Computer Vision

Towards Brain MRI Foundation Models for the Clinic: Findings from the FOMO25 Challenge

Asbjørn Munk, Stefano Cerri, Vardan Nersesjan et al. (+81 more)

Clinical deployment of automated brain MRI analysis faces a fundamental challenge: clinical data is heterogeneous and noisy, and high-quality labels are prohibitively costly to obtain. Self-supervised learning (SSL) can address this by leveraging the vast amounts of unlabeled data produced in clinical workflows to train robust foundation models that adapt out-of-domain with minimal supervision. However, the development of foundation models for brain MRI has been limited by small pretraining datasets and in-domain benchmarking focused on high-quality, research-grade data. To address this gap, we organized the FOMO25 challenge as a satellite event at MICCAI 2025. FOMO25 provided participants with a large pretraining dataset, FOMO60K, and evaluated models on data sourced directly from clinical workflows in few-shot and out-of-domain settings. Tasks covered infarct classification, meningioma segmentation, and brain age regression, and considered both models trained on FOMO60K (method track) and any data (open track). Nineteen foundation models from sixteen teams were evaluated using a standardized containerized pipeline. Results show that (a) self-supervised pretraining improves generalization on clinical data under domain shift, with the strongest models trained out-of-domain surpassing supervised baselines trained in-domain. (b) No single pretraining objective benefits all tasks: MAE favors segmentation, hybrid reconstruction-contrastive objectives favor classification, and (c) strong performance was achieved by small pretrained models, and improvements from scaling model size and training duration did not yield reliable benefits.
Read the full abstract

All authors Asbjørn Munk, Stefano Cerri, Vardan Nersesjan, Christian Hedeager Krag, Jakob Ambsdorf, Pablo Rocamora García, Julia Machnio, Peirong Liu, Suhyun Ahn, Nasrin Akbari, Yasmina Al Khalil, Kimberly Amador, Sina Amirrajab, Tal Arbel, Meritxell Bach Cuadra, Ujjwal Baid, Bhakti Baheti, Jaume Banus, Kamil Barbierik, Christoph Brune, Yansong Bu, Baptiste Callard, Yuhan Chen, Cornelius Crijnen, Corentin Dancette, Peter Drotar, Prasad Dutande, Nils D. Forkert, Saurabh Garg, Jakub Gazda, Matej Gazda, Benoît Gérin, Partha Ghosh, Weikang Gong, Pedro M. Gordaliza, Sam Hashemi, Tobias Heimann, Fucang Jia, Jiexin Jiang, Emily Kaczmarek, Chris Kang, Seung Kwan Kang, Mohammad Khazaei, Julien Khlaut, Petros Koutsouvelis, Jae Sung Lee, Yuchong Li, Mengye Lyu, Mingchen Ma, Anant Madabhushi, Klaus H. Maier-Hein, Pierre Manceron, Andrés Martínez Mora, Moona Mazher, Felix Meister, Nataliia Molchanova, Steven A. Niederer, Leonard Nürnberg, Jinah Park, Abdul Qayyum, Jonas Richiardi, Antoine Saporta, Branislav Setlak, Ning Shen, Justin Szeto, Constantin Ulrich, Puru Vaish, Vibujithan Vigneshwaran, Leroy Volmer, Zihao Wang, Siqi Wei, Anthony Winder, Jelmer M. Wolterink, Maxence Wynen, Chang Yang, Si Young Yie, Mostafa Mehdipour Ghazi, Akshay Pai, Espen Jimenez Solem, Sebastian Nørgaard Llambias, Mikael Boesen, Michael Eriksen Benros, Juan Eugenio Iglesias, Mads Nielsen

Computer Vision

UNIGEOCLIP: Unified Geospatial Contrastive Learning

Guillaume Astruc, Eduard Trulls, Jan Hosang et al. (+2 more)

The growing availability of co-located geospatial data spanning aerial imagery, street-level views, elevation models, text, and geographic coordinates offers a unique opportunity for multimodal representation learning. We introduce UNIGEOCLIP, a massively multimodal contrastive framework to jointly align five complementary geospatial modalities in a single unified embedding space. Unlike prior approaches that fuse modalities or rely on a central pivot representation, our method performs all-to-all contrastive alignment, enabling seamless comparison, retrieval, and reasoning across arbitrary combinations of modalities. We further propose a scaled latitude-longitude encoder that improves spatial representation by capturing multi-scale geographic structure. Extensive experiments across downstream geospatial tasks demonstrate that UNIGEOCLIP consistently outperforms single-modality contrastive models and coordinate-only baselines, highlighting the benefits of holistic multimodal geospatial alignment. A reference implementation is available at https://gastruc.github.io/unigeoclip.
Read the full abstract

All authors Guillaume Astruc, Eduard Trulls, Jan Hosang, Loic Landrieu, Paul-Edouard Sarlin

Computer Vision

GazeVaLM: A Multi-Observer Eye-Tracking Benchmark for Evaluating Clinical Realism in AI-Generated X-Rays

David Wong, Zeynep Isik, Bin Wang et al. (+22 more)

We introduce GazeVaLM, a public eye-tracking dataset for studying clinical perception during chest radiograph authenticity assessment. The dataset comprises 960 gaze recordings from 16 expert radiologists interpreting 30 real and 30 synthetic chest X-rays (generated by diffusion based generative AI) under two conditions: diagnostic assessment and real-fake classification (Visual Turing test). For each image-observer pair, we provide raw gaze samples, fixation maps, scanpaths, saliency density maps, structured diagnostic labels, and authenticity judgments. We extend the protocol to 6 state-of-the-art multimodal LLMs, releasing their predicted diagnoses, authenticity labels, and confidence scores under matched conditions - enabling direct human-AI comparison at both decision and uncertainty levels. We further provide analyses of gaze agreement, inter-observer consistency, and benchmarking of radiologists versus LLMs in diagnostic accuracy and authenticity detection. GazeVaLM supports research in gaze modeling, clinical decision-making, human-AI comparison, generative image realism assessment, and uncertainty quantification. By jointly releasing visual attention data, clinical labels, and model predictions, we aim to facilitate reproducible research on how experts and AI systems perceive, interpret, and evaluate medical images. The dataset is available at https://huggingface.co/datasets/davidcwong/GazeVaLM.
Read the full abstract

All authors David Wong, Zeynep Isik, Bin Wang, Marouane Tliba, Gorkem Durak, Elif Keles, Halil Ertugrul Aktas, Aladine Chetouani, Cagdas Topel, Nicolo Gennaro, Camila Lopes Vendrami, Tugce Agirlar Trabzonlu, Amir Ali Rahsepar, Laetitia Perronne, Matthew Antalek, Onural Ozturk, Gokcan Okur, Andrew C. Gordon, Ayis Pyrros, Frank H. Miller, Amir Borhani, Hatice Savas, Eric Hart, Elizabeth Krupinski, Ulas Bagci

Computer Vision

STS-Mixer: Spatio-Temporal-Spectral Mixer for 4D Point Cloud Video Understanding

Wenhao Li, Xueying Jiang, Gongjie Zhang et al. (+3 more)

4D point cloud videos capture rich spatial and temporal dynamics of scenes which possess unique values in various 4D understanding tasks. However, most existing methods work in the spatiotemporal domain where the underlying geometric characteristics of 4D point cloud videos are hard to capture, leading to degraded representation learning and understanding of 4D point cloud videos. We address the above challenge from a complementary spectral perspective. By transforming 4D point cloud videos into graph spectral signals, we can decompose them into multiple frequency bands each of which captures distinct geometric structures of point cloud videos. Our spectral analysis reveals that the decomposed low-frequency signals capture more coarse shapes while high-frequency signals encode more fine-grained geometry details. Building on these observations, we design Spatio-Temporal-Spectral Mixer (STS-Mixer), a unified framework that mixes spatial, temporal, and spectral representations of point cloud videos. STS-Mixer integrates multi-band delineated spectral signals with spatiotemporal information to capture rich geometries and temporal dynamics, while enabling fine-grained and holistic understanding of 4D point cloud videos. Extensive experiments show that STS-Mixer achieves superior performance consistently across multiple widely adopted benchmarks on both 3D action recognition and 4D semantic segmentation tasks. Code and models are available at https://github.com/Vegetebird/STS-Mixer.
Read the full abstract

All authors Wenhao Li, Xueying Jiang, Gongjie Zhang, Xiaoqin Zhang, Ling Shao, Shijian Lu

Computer Vision

MorphoFlow: Sparse-Supervised Generative Shape Modeling with Adaptive Latent Relevance

Mokshagna Sai Teja Karanam, Tushar Kataria and Shireen Elhabian

Statistical shape modeling (SSM) is central to population level analysis of anatomical variability, yet most existing approaches rely on densely annotated segmentations and fixed latent representations. These requirements limit scalability and reduce flexibility when modeling complex anatomical variation. We introduce MorphoFlow, a sparse supervised generative shape modeling framework that learns compact probabilistic shape representations directly from sparse surface annotations. MorphoFlow integrates neural implicit shape representations with an autodecoder formulation and autoregressive normalizing flows to learn an expressive probabilistic density over the latent shape space. The neural implicit representation enables resolution-agnostic modeling of 3D anatomy, while the autodecoder formulation supports direct optimization of per-instance latent codes under sparse supervision. The autoregressive flow captures the distribution of latent anatomical variability providing a tractable, likelihood-based generative model of shapes. To promote compact and structured latent representations, we incorporate adaptive latent relevance weighting through sparsity-inducing priors, enabling the model to regulate the contribution of individual latent dimensions according to their relevance to the underlying anatomical variation while preserving generative expressivity. The resulting latent space supports uncertainty quantification and anatomically plausible shape synthesis without manual latent dimensionality tuning. Evaluation on publicly available lumbar vertebrae and femur datasets demonstrates accurate high-resolution reconstruction from sparse inputs and recovery of structured modes of anatomical variation consistent with population level trends.
Read the full abstract

Computer Vision

POINTS-Long: Adaptive Dual-Mode Visual Reasoning in MLLMs

Haicheng Wang, Yuan Liu, Yikun Liu et al. (+9 more)

Multimodal Large Language Models (MLLMs) have recently demonstrated remarkable capabilities in cross-modal understanding and generation. However, the rapid growth of visual token sequences–especially in long-video and streaming scenarios–poses a major challenge to their scalability and real-world deployment. Thus, we introduce POINTS-Long, a native dual-mode MLLM featuring dynamic visual token scaling inspired by the human visual system. The model supports two complementary perception modes: focus mode and standby mode, enabling users to dynamically trade off efficiency and accuracy during inference. On fine-grained visual tasks, the focus mode retains the optimal performance, while on long-form general visual understanding, the standby mode retains 97.7-99.7% of the original accuracy using only 1/40-1/10th of the visual tokens. Moreover, POINTS-Long natively supports streaming visual understanding via a dynamically detachable KV-cache design, allowing efficient maintenance of ultra-long visual memory. Our work provides new insights into the design of future MLLMs and lays the foundation for adaptive and efficient long-form visual understanding.
Read the full abstract

All authors Haicheng Wang, Yuan Liu, Yikun Liu, Zhemeng Yu, Zhongyin Zhao, Yangxiu You, Zilin Yu, Le Tian, Xiao Zhou, Jie Zhou, Weidi Xie, Yanfeng Wang

Computer Vision

Geoparsing: Diagram Parsing for Plane and Solid Geometry with a Unified Formal Language

Peijie Wang, Ming-Liang Zhang, Jun Cao et al. (+10 more)

Multimodal Large Language Models (MLLMs) have achieved remarkable progress but continue to struggle with geometric reasoning, primarily due to the perception bottleneck regarding fine-grained visual elements. While formal languages have aided plane geometry understanding, solid geometry which requires spatial understanding remains largely unexplored. In this paper, we address this challenge by designing a unified formal language that integrates plane and solid geometry, comprehensively covering geometric structures and semantic relations. We construct GDP-29K, a large-scale dataset comprising 20k plane and 9k solid geometry samples collected from diverse real-world sources, each paired with its ground-truth formal description. To ensure syntactic correctness and geometric consistency, we propose a training paradigm that combines Supervised Fine-Tuning with Reinforcement Learning via Verifiable Rewards. Experiments show that our approach achieves state-of-the-art parsing performance. Furthermore, we demonstrate that our parsed formal descriptions serve as a critical cognitive scaffold, significantly boosting MLLMs' capabilities for downstream geometry reasoning tasks. Our data and code are available at Geoparsing.
Read the full abstract

All authors Peijie Wang, Ming-Liang Zhang, Jun Cao, Chao Deng, Dekang Ran, Hongda Sun, Pi Bu, Xuan Zhang, Yingyao Wang, Jun Song, Bo Zheng, Fei Yin, Cheng-Lin Liu

Computer Vision

Learning Robustness at Test-Time from a Non-Robust Teacher

Stefano Bianchettin, Giulio Rossolini and Giorgio Buttazzo

Nowadays, pretrained models are increasingly used as general-purpose backbones and adapted at test-time to downstream environments where target data are scarce and unlabeled. While this paradigm has proven effective for improving clean accuracy on the target domain, adversarial robustness has received far less attention, especially when the original pretrained model is not explicitly designed to be robust. This raises a practical question: can a pretrained, non-robust model be adapted at test-time to improve adversarial robustness on a target distribution? To face this question, this work studies how adversarial training strategies behave when integrated into adaptation schemes for the unsupervised test-time setting, where only a small set of unlabeled target samples is available. It first analyzes how classical adversarial training formulations can be extended to this scenario, showing that straightforward distillation-based adaptations remain unstable and highly sensitive to hyperparameter tuning, particularly when the teacher itself is non-robust. To address these limitations, the work proposes a label-free framework that uses the predictions of a non-robust teacher model as a semantic anchor for both the clean and adversarial objectives during adaptation. We further provide theoretical insights showing that our formulation yields a more stable alternative to the self-consistency-based regularization commonly used in classical adversarial training. Experiments evaluate the proposed approach on CIFAR-10 and ImageNet under induced photometric transformations. The results support the theoretical insights by showing that the proposed approach achieves improved optimization stability, lower sensitivity to parameter choices, and a better robustness-accuracy trade-off than existing baselines in this post-deployment test-time setting.
Read the full abstract

Computer Vision

MLLM-as-a-Judge Exhibits Model Preference Bias

Shuitsu Koyama, Yuiga Wada, Daichi Yashima et al. (+1 more)

Automatic evaluation using multimodal large language models (MLLMs), commonly referred to as MLLM-as-a-Judge, has been widely used to measure model performance. If such MLLM-as-a-Judge methods were biased, they could distort model comparisons and benchmark-driven scientific progress. However, it remains unclear to what extent MLLM-as-a-Judge methods favor or disfavor text generated by specific MLLMs. In this study, we propose Philautia-Eval to investigate such model-specific preference bias. Philautia-Eval quantifies the degree of the bias by disentangling preference tendencies from differences in generation quality. Using 1.29M caption-score pairs collected from 12 MLLMs, we found that representative MLLMs tend to exhibit self-preference bias. Moreover, experimental results indicate mutual preference bias within particular model families, which is potentially driven by reused connectors and overlapping instruction-tuning resources. Finally, we introduce a simple ensemble of MLLMs, Pomms. Our results demonstrated that Pomms effectively mitigated the model-specific preference bias while maintaining performance.
Read the full abstract

All authors Shuitsu Koyama, Yuiga Wada, Daichi Yashima, Komei Sugiura

Computer Vision

GeomPrompt: Geometric Prompt Learning for RGB-D Semantic Segmentation Under Missing and Degraded Depth

Krishna Jaganathan and Patricio Vela

Multimodal perception systems for robotics and embodied AI often assume reliable RGB-D sensing, but in practice, depth is frequently missing, noisy, or corrupted. We thus present GeomPrompt, a lightweight cross-modal adaptation module that synthesizes a task-driven geometric prompt from RGB alone for the fourth channel of a frozen RGB-D semantic segmentation model, without depth supervision. We further introduce GeomPrompt-Recovery, an adaptation module that compensates for degraded depth by predicting the fourth channel correction relevant for the frozen segmenter. Both modules are trained solely with downstream segmentation supervision, enabling recovery of the geometric prior useful for segmentation, rather than estimating depth signals. On SUN RGB-D, GeomPrompt improves over RGB-only inference by +6.1 mIoU on DFormer and +3.0 mIoU on GeminiFusion, while remaining competitive with strong monocular depth estimators. For degraded depth, GeomPrompt-Recovery consistently improves robustness, yielding gains up to +3.6 mIoU under severe depth corruptions. GeomPrompt is also substantially more efficient than monocular depth baselines, reaching 7.8 ms latency versus 38.3 ms and 71.9 ms. These results suggest that task-driven geometric prompting is an efficient mechanism for cross-modal compensation under missing and degraded depth inputs in RGB-D perception.
Read the full abstract

Computer Vision

Seeing Through Touch: Tactile-Driven Visual Localization of Material Regions

Seongyu Kim, Seungwoo Lee, Hyeonggon Ryu et al. (+2 more)

We address the problem of tactile localization, where the goal is to identify image regions that share the same material properties as a tactile input. Existing visuo-tactile methods rely on global alignment and thus fail to capture the fine-grained local correspondences required for this task. The challenge is amplified by existing datasets, which predominantly contain close-up, low-diversity images. We propose a model that learns local visuo-tactile alignment via dense cross-modal feature interactions, producing tactile saliency maps for touch-conditioned material segmentation. To overcome dataset constraints, we introduce: (i) in-the-wild multi-material scene images that expand visual diversity, and (ii) a material-diversity pairing strategy that aligns each tactile sample with visually varied yet tactilely consistent images, improving contextual localization and robustness to weak signals. We also construct two new tactile-grounded material segmentation datasets for quantitative evaluation. Experiments on both new and existing benchmarks show that our approach substantially outperforms prior visuo-tactile methods in tactile localization.
Read the full abstract

All authors Seongyu Kim, Seungwoo Lee, Hyeonggon Ryu, Joon Son Chung, Arda Senocak

Computer Vision

Finetune Like You Pretrain: Boosting Zero-shot Adversarial Robustness in Vision-language Models

Songlong Xing, Weijie Wang, Zhengyu Zhao et al. (+3 more)

Despite their impressive zero-shot abilities, vision-language models such as CLIP have been shown to be susceptible to adversarial attacks. To enhance its adversarial robustness, recent studies finetune the pretrained vision encoder of CLIP with adversarial examples on a proxy dataset such as ImageNet by aligning adversarial images with correct class labels. However, these methods overlook the important roles of training data distributions and learning objectives, resulting in reduced zero-shot capabilities and limited transferability of robustness across domains and datasets. In this work, we propose a simple yet effective paradigm AdvFLYP, which follows the training recipe of CLIP's pretraining process when performing adversarial finetuning to the model. Specifically, AdvFLYP finetunes CLIP with adversarial images created based on image-text pairs collected from the web, and match them with their corresponding texts via a contrastive loss. To alleviate distortion of adversarial image embeddings of noisy web images, we further propose to regularise AdvFLYP by penalising deviation of adversarial image features. We show that logit- and feature-level regularisation terms benefit robustness and clean accuracy, respectively. Extensive experiments on 14 downstream datasets spanning various domains show the superiority of our paradigm over mainstream practices. Our code and model weights are released at https://github.com/Sxing2/AdvFLYP.
Read the full abstract

All authors Songlong Xing, Weijie Wang, Zhengyu Zhao, Jindong Gu, Philip Torr, Nicu Sebe

Computer Vision

Training-Free Model Ensemble for Single-Image Super-Resolution via Strong-Branch Compensation

Gengjia Chang, Xining Ge, Weijun Yuan et al. (+4 more)

Single-image super-resolution has progressed from deep convolutional baselines to stronger Transformer and state-space architectures, yet the corresponding performance gains typically come with higher training cost, longer engineering iteration, and heavier deployment burden. In many practical settings, multiple pretrained models with partially complementary behaviors are already available, and the binding constraint is no longer architectural capacity but how effectively their outputs can be combined without additional training. Rather than pursuing further architectural redesign, this paper proposes a training-free output-level ensemble framework. A dual-branch pipeline is constructed in which a Hybrid attention network with TLC inference provides stable main reconstruction, while a MambaIRv2 branch with geometric self-ensemble supplies strong compensation for high-frequency detail recovery. The two branches process the same low-resolution input independently and are fused in the image space via a lightweight weighted combination, without updating any model parameters or introducing an additional trainable module. As our solution to the NTIRE 2026 Image Super-Resolution ($\times 4$) Challenge, the proposed design consistently improves over the base branch and slightly exceeds the pure strong branch in PSNR at the best operating point under a unified DIV2K bicubic $\times 4$ evaluation protocol. Ablation studies confirm that output-level compensation provides a low-overhead and practically accessible upgrade path for existing super-resolution systems.
Read the full abstract

All authors Gengjia Chang, Xining Ge, Weijun Yuan, Zhan Li, Qiurong Song, Luen Zhu, Shuhong Liu

Computer Vision

The Impact of Federated Learning on Distributed Remote Sensing Archives

Anand Umashankar, Karam Tomotaki-Dawoud and Nicolai Schneider

Remote sensing archives are inherently distributed: Earth observation missions such as Sentinel-1, Sentinel-2, and Sentinel-3 have collectively accumulated more than 5 petabytes of imagery, stored and processed across many geographically dispersed platforms. Training machine learning models on such data in a centralized fashion is impractical due to data volume, sovereignty constraints, and geographic distribution. Federated learning (FL) addresses this by keeping data local and exchanging only model updates. A central challenge for remote sensing is the non-IID nature of Earth observation data: label distributions vary strongly by geographic region, degrading the convergence of standard FL algorithms. In this paper, we conduct a systematic empirical study of three FL strategies – FedAvg, FedProx, and bulk synchronous parallel (BSP) – applied to multi-label remote sensing image classification under controlled non-IID label-skew conditions. We evaluate three convolutional neural network (CNN) architectures of increasing depth (LeNet, AlexNet, and ResNet-34) and analyze the joint effect of algorithm choice, model capacity, client fraction, client count, batch size, and communication cost. Experiments on the UC Merced multi-label dataset show that FedProx outperforms FedAvg for deeper architectures under data heterogeneity, that BSP approaches centralized accuracy at the cost of high sequential communication, and that LeNet provides the best accuracy-communication trade-off for the dataset scale considered.
Read the full abstract

Computer Vision

Progressively Texture-Aware Diffusion for Contrast-Enhanced Sparse-View CT

Tianqi Wang, Wenchao Du and Hongyu Yang

Diffusion-based sparse-view CT (SVCT) imaging has achieved remarkable advancements in recent years, thanks to its more stable generative capability. However, recovering reliable image content and visually consistent textures is still a crucial challenge. In this paper, we present a Progressively Texture-aware Diffusion (PTD) model, a coarse-to-fine learning framework tailored for SVCT. Specifically, PTD comprises a basic reconstructive module PTD$_{\textit{rec}}$ and a conditional diffusion module PTD$_{\textit{diff}}$. PTD$_{\textit{rec}}$ first learns a deterministic mapping to recover the majority of the underlying low-frequency signals (i.e., coarse content with smoothed textures), which serves as the initial estimation to enable fidelity. Moreover, PTD$_{\textit{diff}}$ aims to reconstruct high-fidelity details for coarse prediction, which explores a dual-domain guided conditional diffusion to generate reliable and consistent textures. Extensive experiments on sparse-view CT reconstruction demonstrate that our PTD achieves superior performance in terms of structure similarity and visual appeal with only a few sampling steps, which mitigates the randomness inherent in general diffusion models and enables a better trade-off between visual quality and fidelity of high-frequency details.
Read the full abstract

Computer Vision

CLAY: Conditional Visual Similarity Modulation in Vision-Language Embedding Space

Sohwi Lim, Lee Hyoseok, Jungjoon Park et al. (+1 more)

Human perception of visual similarity is inherently adaptive and subjective, depending on the users' interests and focus. However, most image retrieval systems fail to reflect this flexibility, relying on a fixed, monolithic metric that cannot incorporate multiple conditions simultaneously. To address this, we propose CLAY, an adaptive similarity computation method that reframes the embedding space of pretrained Vision-Language Models (VLMs) as a text-conditional similarity space without additional training. This design separates the textual conditioning process and visual feature extraction, allowing highly efficient and multi-conditioned retrieval with fixed visual embeddings. We also construct a synthetic evaluation dataset CLAY-EVAL, for comprehensive assessment under diverse conditioned retrieval settings. Experiments on standard datasets and our proposed dataset show that CLAY achieves high retrieval accuracy and notable computational efficiency compared to previous works.
Read the full abstract

All authors Sohwi Lim, Lee Hyoseok, Jungjoon Park, Tae-Hyun Oh

Computer Vision

Beyond Attention Scores: SVD-Based Vision Token Pruning for Efficient Vision-Language Models

Yvon Apedo, Martyna Poreba, Michal Szczepanski et al. (+1 more)

Vision-Language Models (VLMs) have revolutionized multi-modal learning by jointly processing visual and textual information. Yet, they face significant challenges due to the high computational and memory demands of processing long sequences of vision tokens. Many existing methods rely on local heuristics, such as attention scores or token norms. However, these criteria suffer from positional bias and information dispersion, limiting their ability to preserve essential content at high pruning ratios and leading to performance degradation on visually detailed images. To address these issues, we propose SVD-Prune, a training-free, plug-and-play token pruning method based on Singular Value Decomposition. It decomposes the vision token feature matrix and selects the top-k tokens using statistical leverage scores, ensuring only tokens contributing most to the dominant global variance are preserved. Experiments show that SVD-Prune consistently outperforms prior pruning methods under extreme vision token budgets, maintaining strong performance even with 32 and 16 vision tokens.
Read the full abstract

All authors Yvon Apedo, Martyna Poreba, Michal Szczepanski, Samia Bouchafa

Computer Vision · Cross-listed from cs.LG

Continuous Adversarial Flow Models

Shanchuan Lin, Ceyuan Yang, Zhijie Lin et al. (+2 more)

We propose continuous adversarial flow models, a type of continuous-time flow model trained with an adversarial objective. Unlike flow matching, which uses a fixed mean-squared-error criterion, our approach introduces a learned discriminator to guide training. This change in objective induces a different generalized distribution, which empirically produces samples that are better aligned with the target data distribution. Our method is primarily proposed for post-training existing flow-matching models, although it can also train models from scratch. On the ImageNet 256px generation task, our post-training substantially improves the guidance-free FID of latent-space SiT from 8.26 to 3.63 and of pixel-space JiT from 7.17 to 3.57. It also improves guided generation, reducing FID from 2.06 to 1.53 for SiT and from 1.86 to 1.80 for JiT. We further evaluate our approach on text-to-image generation, where it achieves improved results on both the GenEval and DPG benchmarks.
Read the full abstract

All authors Shanchuan Lin, Ceyuan Yang, Zhijie Lin, Hao Chen, Haoqi Fan

Computer Vision

TAG-Head: Time-Aligned Graph Head for Plug-and-Play Fine-grained Action Recognition

Imtiaz Ul Hassan, Nik Bessis and Ardhendu Behera

Fine-grained human action recognition (FHAR) is challenging because visually similar actions differ by subtle spatio-temporal cues. Many recent systems enhance discriminability with extra modalities (e.g., pose, text, optical flow), but this increases annotation burden and computational cost. We introduce TAG-Head, a lightweight spatio-temporal graph head that upgrades standard 3D backbones (SlowFast, R(2+1)D-34, I3D, etc.) for FHAR using RGB only. Our pipeline first applies a Transformer encoder with learnable 3D positional encodings to the backbone tokens, capturing long-range dependencies across space and time. The resulting features are then refined by a graph in which (i) fully-connected intra-frame edges to resolve subtle appearance differences within frames, and (ii) time-aligned temporal edges that connect features at the same spatial location across frames to stabilise motion cues without over-smoothing. The head is compact (little parameter/FLOP overhead), plug-and-play across backbones, and trained end-to-end with the backbone. Extensive evaluations on FineGym (Gym99 and Gym288) and HAA500 show that TAG-Head sets a new state-of-the-art among RGB-only models and surpasses many recent multimodal approaches (video + pose + text) that rely on privileged information. Ablations disentangle the contributions of the Transformer and the graph topology, and complexity analyses confirm low latency. TAG-Head advances FHAR by explicitly coupling global context with high-resolution spatial interactions and low-variance temporal continuity inside a slim, composable graph head. The simplicity of the design enables straightforward adoption in practical systems that favour RGB-only sensors, while delivering performance gains typically associated with heavier or multimodal models. Code will be released on GitHub.
Read the full abstract

Computer Vision

Revisiting Compositionality in Dual-Encoder Vision-Language Models: The Role of Inference

Imanol Miranda, Ander Salaberria, Eneko Agirre et al. (+1 more)

Dual-encoder Vision-Language Models (VLMs) such as CLIP are often characterized as bag-of-words systems due to their poor performance on compositional benchmarks. We argue that this limitation may stem less from deficient representations than from the standard inference protocol based on global cosine similarity. First, through controlled diagnostic experiments, we show that explicitly enforcing fine-grained region-segment alignment at inference dramatically improves compositional performance without updating pretrained encoders. We then introduce a lightweight transformer that learns such alignments directly from frozen patch and token embeddings. Comparing against full fine-tuning and prior end-to-end compositional training methods, we find that although these approaches improve in-domain retrieval, their gains do not consistently transfer under distribution shift. In contrast, learning localized alignment over frozen representations matches full fine-tuning on in-domain retrieval while yielding substantial improvements on controlled out-of-domain compositional benchmarks. These results identify global embedding matching as a key bottleneck in dual-encoder VLMs and highlight the importance of alignment mechanisms for robust compositional generalization.
Read the full abstract

All authors Imanol Miranda, Ander Salaberria, Eneko Agirre, Gorka Azkune

Computer Vision · Cross-listed from cs.AI

Anthropogenic Regional Adaptation in Multimodal Vision-Language Model

Samuel Cahyawijaya, Peerat Limkonchotiwat, Tack Hwa Wong et al. (+45 more)

While the field of vision-language (VL) has achieved remarkable success in integrating visual and textual information across multiple languages and domains, there is still no dedicated framework for assessing human-centric alignment in vision-language systems. We offer two contributions to address this gap. First, we introduce Anthropogenic Regional Adaptation: a novel paradigm that aims to optimize model relevance to specific regional contexts while ensuring the retention of global generalization capabilities. Second, we present a simple, but effective adaptation method named Geographical-generalization-made-easy (GG-EZ), which utilizes regional data filtering and model merging. Through comprehensive experiments on 3 VL architectures: large vision-language models, text-to-image diffusion models, and vision-language embedding models, and a case study in Southeast Asia (SEA) regional adaptation, we demonstrate the importance of Anthropogenic Regional Adaptation and the effectiveness of GG-EZ, showing 5-15% gains in cultural relevance metrics across SEA while maintaining over 98% of global performance and even occasionally surpassing it. Our findings establish Anthropogenic Regional Alignment as a foundational paradigm towards applicability of multimodal vision-language models in diverse regions and demonstrate a simple-yet-effective baseline method that optimizes regional value alignment while preserving global generalization.
Read the full abstract

All authors Samuel Cahyawijaya, Peerat Limkonchotiwat, Tack Hwa Wong, Hitesh Laxmichand Patel, Amit Agarwal, Manuel Antonio Rufino, Carlos Rafael Catalan, Muhammad Reza Qorib, Vicky Feliren, Holy Lovenia, Aye Hninn Khine, Frederikus Hudi, David Anugraha, Alham Fikri Aji, Romrawin Chumpu, Viet-Thanh Pham, Minghan Wang, Mohamed Fazli Imam, Ruochen Zhang, Joseph Marvin Imperial, Khumaisa Nur'aini, Do Xuan Long, Musa Izzanardi Wijanarko, Joel Ruben Antony Moniz, Patrick Amadeus Irawan, Hanif Muhammad Zhafran, Isaiah Flores, Salsabila Zahirah Pranida, Jun Kevin, Jostin Jerico Rosal, Patricia Nicole Monderin, Kun Kerdthaisong, Ahmad Mustafid, My Chiffon Nguyen, Natchapon Jongwiriyanurak, Siva Worajitwannakul, Haochen Li, Adrian Xuan Wei Lim, Bin Wang, Muhammad Ravi Shulthan Habibi, Lynnette Hui Xian Ng, Mithil Bangera, Yeshil Bangera, Priyaranjan Pattnayak, Dun Li Chan, Sherissa Caren Djuniwar, Cho Chan Myei Oo, Hee Ming Shan

Computer Vision

NTIRE 2026 Challenge on Robust AI-Generated Image Detection in the Wild

Aleksandr Gushchin, Khaled Abud, Ekaterina Shumitskaya et al. (+51 more)

This paper presents an overview of the NTIRE 2026 Challenge on Robust AI-Generated Image Detection in the Wild, held in conjunction with the NTIRE workshop at CVPR 2026. The goal of this challenge was to develop detection models capable of distinguishing real images from generated ones in realistic scenarios: the images are often transformed (cropped, resized, compressed, blurred) for practical usage, and therefore, the detection models should be robust to such transformations. The challenge is based on a novel dataset consisting of 108,750 real and 185,750 AI-generated images from 42 generators comprising a large variety of open-source and closed-source models of various architectures, augmented with 36 image transformations. Methods were evaluated using ROC AUC on the full test set, including both transformed and untransformed images. A total of 511 participants registered, with 20 teams submitting valid final solutions. This report provides a comprehensive overview of the challenge, describes the proposed solutions, and can be used as a valuable reference for researchers and practitioners in increasing the robustness of the detection models to real-world transformations.
Read the full abstract

All authors Aleksandr Gushchin, Khaled Abud, Ekaterina Shumitskaya, Artem Filippov, Georgii Bychkov, Sergey Lavrushkin, Mikhail Erofeev, Anastasia Antsiferova, Changsheng Chen, Shunquan Tan, Radu Timofte, Dmitry Vatolin, Chuanbiao Song, Zijian Yu, Hao Tan, Jun Lan, Zhiqiang Yang, Yongwei Tang, Zhiqiang Wu, Jia Wen Seow, Hong Vin Koay, Haodong Ren, Feng Xu, Shuai Chen, Ruiyang Xia, Qi Zhang, Yaowen Xu, Zhaofan Zou, Hao Sun, Dagong Lu, Mufeng Yao, Xinlei Xu, Fei Wu, Fengjun Guo, Cong Luo, Hardik Sharma, Aashish Negi, Prateek Shaily, Jayant Kumar, Sachin Chaudhary, Akshay Dudhane, Praful Hambarde, Amit Shukla, Zhilin Tu, Fengpeng Li, Jiamin Zhang, Jianwei Fei, Kemou Li, Haiwei Wu, Bilel Benjdira, Anas M. Ali, Wadii Boulila, Chenfan Qu, Junchi Li

Computer Vision

PACO: Proxy-Task Alignment and Online Calibration for On-the-Fly Category Discovery

Weidong Tang, Bohan Zhang, Zhixiang Chi et al. (+3 more)

On-the-Fly Category Discovery (OCD) requires a model, trained on an offline support set, to recognize known classes while discovering new ones from an online streaming sequence. Existing methods focus heavily on offline training. They aim to learn discriminative representations on the support set so that novel classes can be separated at test time. However, their discovery mechanism at inference is typically reduced to a single threshold. We argue that this paradigm is fundamentally flawed as OCD is not a static classification problem, but a dynamic process. The model must continuously decide 1) whether a sample belongs to a known class, 2) matches an existing novel category, or 3) should initiate a new one. Moreover, prior methods treat the support set as fixed knowledge. They do not update their decision boundaries as new evidence arrives during inference. This leads to unstable and inconsistent category formation. Our experiments confirm these issues. With properly calibrated and adaptive thresholds, substantial improvements can be achieved, even without changing the representation. Motivated by this, we propose PACO, a support-set-calibrated, tree-structured online decision framework. The framework models inference as a sequence of hierarchical decisions, including known-class routing, birth-aware novel assignment, and attach-versus-create operations over a dynamic prototype memory. Furthermore, we simulate the proxy discovery process to initialize the thresholds during offline training to align with inference. Thresholds are continuously updated during inference using mature novel prototypes. Importantly, PACO requires no heavy training and no dataset-specific tuning. It can be directly integrated into existing OCD pipelines as an inference-time module. Extensive experiments show significant improvements over SOTA baselines across seven benchmarks.
Read the full abstract

All authors Weidong Tang, Bohan Zhang, Zhixiang Chi, ZiZhang Wu, Yang Wang, Yanan Wu

Computer Vision

Degradation-Aware and Structure-Preserving Diffusion for Real-World Image Super-Resolution

Yang Ji, Zonghao Chen, Zhihao Xue et al. (+1 more)

Real-world image super-resolution is particularly challenging for diffusion models because real degradations are complex, heterogeneous, and rarely modeled explicitly. We propose a degradation-aware and structure-preserving diffusion framework for real-world SR. Specifically, we introduce Degradation-aware Token Injection, which encodes lightweight degradation statistics from low-resolution inputs and fuses them with semantic conditioning features, enabling explicit degradation-aware restoration. We further propose Spatially Asymmetric Noise Injection, which modulates diffusion noise with local edge strength to better preserve structural regions during training. Both modules are lightweight add-ons to the adopted diffusion SR framework, requiring only minor modifications to the conditioning pipeline. Experiments on DIV2K and RealSR show that our method delivers competitive no-reference perceptual quality and visually more realistic restoration results than recent baselines, while maintaining a favorable perception–distortion trade-off. Ablations confirm the effectiveness of each module and their complementary gains when combined. The code and model are publicly available at https://github.com/jiyang0315/DASP-SR.git.
Read the full abstract

All authors Yang Ji, Zonghao Chen, Zhihao Xue, Junqin Hu

Computer Vision

Beyond Model Design: Data-Centric Training and Self-Ensemble for Gaussian Color Image Denoising

Gengjia Chang, Xining Ge, Weijun Yuan et al. (+4 more)

This paper presents our solution to the NTIRE 2026 Image Denoising Challenge (Gaussian color image denoising at fixed noise level $\sigma = 50$). Rather than proposing a new restoration backbone, we revisit the performance boundary of the mature Restormer architecture from two complementary directions: stronger data-centric training and more complete Test-Time capability release. Starting from the public Restormer $\sigma\!=\!50$ baseline, we expand the standard multi-dataset training recipe with larger and more diverse public image corpora and organize optimization into two stages. At inference, we apply $\times 8$ geometric self-ensemble to further release model capacity. A TLC-style local inference wrapper is retained for implementation consistency; however, systematic ablation reveals its quantitative contribution to be negligible in this setting. On the challenge validation set of 100 images, our final submission achieves 30.762 dB PSNR and 0.861 SSIM, improving over the public Restormer $\sigma\!=\!50$ pretrained baseline by up to 3.366 dB PSNR. Ablation studies show that the dominant gain originates from the expanded training corpus and the two-stage optimization schedule, and self-ensemble provides marginal but consistent improvement.
Read the full abstract

All authors Gengjia Chang, Xining Ge, Weijun Yuan, Zhan Li, Qiurong Song, Luen Zhu, Shuhong Liu

Computer Vision

HuiYanEarth-SAR: A Foundation Model for High-Fidelity and Low-Cost Global Remote Sensing Imagery Generation

Yongxiang Liu, Jie Zhou, Yafei Song et al. (+2 more)

Synthetic Aperture Radar (SAR) imagery generation is essential for deepening the study of scattering mechanisms, establishing trustworthy electromagnetic scene models, and fundamentally alleviating the data scarcity bottleneck that constrains development in this field. However, existing methods find it difficult to simultaneously ensure high fidelity in both global geospatial semantics and microscopic scattering mechanisms, resulting in severe challenges for global generation. To address this, we propose HuiYanEarth-SAR, the first foundational SAR imagery generation model based on AlphaEarth and integrated scattering mechanisms. By injecting geospatial priors to control macroscopic structures and utilizing implicit scattering characteristic modeling to ensure the authenticity of microscopic textures, we achieve the capability of generating high-fidelity SAR images for global locations solely based on geographic coordinates. This study not only constructs an efficient SAR scene simulator but also establishes a bridge connecting geography, scatter mechanism, and artificial intelligence from a methodological standpoint. It advances SAR research by expanding the paradigm from perception and understanding to simulation and creation, providing key technical support for constructing a high-confidence digital twin of the Earth.
Read the full abstract

All authors Yongxiang Liu, Jie Zhou, Yafei Song, Tianpeng Liu, Li Liu

Computer Vision

Observe Less, Understand More: Cost-aware Cross-scale Observation for Remote Sensing Understanding

Zhenghao Xie, Jing Xiao, Zhenqi Wang et al. (+4 more)

Remote sensing understanding inherently requires multi-resolution observation, since different targets and application tasks demand different levels of spatial detail. While low-resolution (LR) imagery enables efficient global observation, high-resolution (HR) imagery provides critical local details at a much higher acquisition cost and with limited coverage. This motivates a cross-scale sensing strategy that selectively acquires HR imagery guided by LR-based global perception to improve task performance under constrained cost. Existing HR sampling methods typically make selection decisions from isolated LR patches, thereby ignoring fine-grained intra-patch importance and cross-patch contextual interactions, leading to fragmented feature representation and suboptimal scene reasoning under sparse HR observations. To address this issue, we formulate cross-scale remote sensing understanding as a unified cost-aware problem that couples fine-grained HR sampling with cross-patch representation prediction, enabling more effective task reasoning with fewer HR observations. Furthermore, we present GL-10M, a high- and low-resolution dataset with nearly 100,000 scene pairs and 10 million images for large-scale cross-resolution pretraining. Extensive experiments on recognition and retrieval tasks show that our method consistently achieves a superior performance-cost trade-off. The code is publicly available at https://github.com/xzhacc/CrossSO.
Read the full abstract

All authors Zhenghao Xie, Jing Xiao, Zhenqi Wang, Kexin Ma, Liang Liao, Gui-song Xia, Mi Wang

Computer Vision

Online Reasoning Video Object Segmentation

Jinyuan Liu, Yang Wang, Zeyu Zhao et al. (+3 more)

Reasoning video object segmentation predicts pixel-level masks in videos from natural-language queries that may involve implicit and temporally grounded references. However, existing methods are developed and evaluated in an offline regime, where the entire video is available at inference time and future frames can be exploited for retrospective disambiguation, deviating from real-world deployments that require strictly causal, frame-by-frame decisions. We study Online Reasoning Video Object Segmentation (ORVOS), where models must incrementally interpret queries using only past and current frames without revisiting previous predictions, while handling referent shifts as events unfold. To support evaluation, we introduce ORVOSB, a benchmark with frame-level causal annotations and referent-shift labels, comprising 210 videos, 12,907 annotated frames, and 512 queries across five reasoning categories. We further propose a baseline with continually-updated segmentation prompts and a structured temporal token reservoir for long-horizon reasoning under bounded computation. Experiments show that existing methods struggle under strict causality and referent shifts, while our baseline establishes a strong foundation for future research.
Read the full abstract

All authors Jinyuan Liu, Yang Wang, Zeyu Zhao, Weixin Li, Song Wang, Ruize Han

Computer Vision

SCD4VPR: Multi-modal Scene Change Detection for Long-term Visual Place Recognition Database Update

Diwei Sheng, Vijayraj Gohil, Satyam Gaba et al. (+5 more)

Long-term autonomy in mobile robotics requires maps that remain accurate as environments change over time. Visual Place Recognition (VPR), a core localization capability, degrades sharply as the temporal gap between query and database images grows, particularly across seasonal transitions. Scene Change Detection (SCD) offers a principled mechanism for database maintenance, but existing methods rely on binary, uni-modal visual features that cannot distinguish structural changes from viewpoint-induced differences - a distinction essential for correct update decisions. We propose SCD4VPR, a scene change detection that jointly reasons about what has changed and distinguishes genuine change from viewpoint-induced difference in a unified vision-language framework. SCD4VPR fuses VLM-generated semantic descriptions with visual features via cross-modal attention and refines predictions with geometric-semantic matching, producing multi-class change masks that separately identify object changes, appearance changes, and viewpoint-induced changes. We introduce NYC-CD, the first real-world street-view SCD benchmark with pixel-level multi-class annotations across 8,122 image pairs. Experiments across four SCD benchmarks show that SCD4VPR consistently improves three architecturally distinct backbones. In a controlled VPR database maintenance experiment on NYU-VPR spanning summer through late winter, we confirm that retrieval performance deteriorates substantially when the database is left unchanged, and show that SCD4VPR-guided updates recover most of this loss (+30.1 R@1 at the largest time gap) while keeping the database far more compact than naive append.
Read the full abstract

All authors Diwei Sheng, Vijayraj Gohil, Satyam Gaba, Zihan Liu, Giles Hamilton-Fletcher, John-Ross Rizzo, Yongqing Liang, Chen Feng

Computer Vision

GS4City: Hierarchical Semantic Gaussian Splatting via City-Model Priors

Qilin Zhang, Jinyu Zhu, Olaf Wysocki et al. (+2 more)

Recent semantic 3D Gaussian Splatting (3DGS) methods primarily rely on 2D foundation models, often yielding ambiguous boundaries and limited support for structured urban semantics. While city models such as CityGML encode hierarchically organized semantics together with building geometry, these labels cannot be directly mapped to Gaussian primitives. We present GS4City, a hierarchical semantic Gaussian Splatting method that incorporates city-model priors for urban scene understanding. GS4City derives reliable image-aligned masks from Level of Detail (LoD) 3 CityGML models via two-pass raycasting, explicitly using parent-child relations to validate and recover fine-grained facade elements. It then fuses these geometry-grounded masks with foundation-model predictions to establish scene-consistent instance correspondences, and learns a compact identity encoding for each Gaussian under joint 2D identity supervision and 3D spatial regularization. Experiments on the TUM2TWIN and Gold Coast datasets show that GS4City effectively incorporates structured building semantics into Gaussian scene representations, outperforming existing 2D-driven semantic 3DGS baselines, including LangSplat and Gaga, by up to 15.8 IoU points in coarse building segmentation and 14.2 mIoU points in fine-grained semantic segmentation. By bridging structured city models and photorealistic Gaussian scene representations, GS4City enables semantically queryable and structure-aware urban reconstruction. Code is available at https://github.com/Jinyzzz/GS4City.
Read the full abstract

All authors Qilin Zhang, Jinyu Zhu, Olaf Wysocki, Benjamin Busam, Boris Jutzi