WACV 2026 论文列表

Wildfire smoke is transparent, amorphous, and often visually confounded with clouds, making early-stage detection particularly challenging. In this work, we introduce a benchmark, called SmokeBench, to evaluate the ability of multimodal large language models (MLLMs) to recognize and localize wildfire smoke in images. The benchmark consists of four tasks: (1) smoke classification, (2) tile-based smoke localization, (3) grid-based smoke localization, and (4) smoke detection. We evaluate several MLLMs, including Idefics2, Qwen2.5-VL, InternVL3, Unified-IO 2, Grounding DINO, GPT-4o, and Gemini-2.5 Pro. Our results show that while some models can classify the presence of smoke when it covers a large area, all models struggle with accurate localization, especially in the early stages. Further analysis reveals that smoke volume is strongly correlated with model performance, whereas contrast plays a comparatively minor role. These findings highlight critical limitations of current MLLMs for safety-critical wildfire monitoring and underscore the need for methods that improve early-stage smoke localization.

Despite rapid advances, text-to-image (T2I) models still falter in generating anatomically coherent and semantically grounded humans. We introduce HumanBench, a large-scale (35K-image), privacy-friendly benchmark that rigorously evaluates T2I models across four axes: template consistency, spatial reasoning, action understanding, and texture recognition. To quantify alignment, we propose two novel metrics--Agreement and Distinction--capturing both fidelity to prompts and semantic contrast with counterfactuals and negations.Evaluating six leading models, we uncover persistent failures including disfigurements, species leakage, texture-object mismatches, and counting errors, especially under compound prompts. A complementary human study reveals that image realism and correctness degrade with prompt complexity, validating our automated assessments. HumanBench offers the first comprehensive audit of human-centric T2I generation, setting a new standard for benchmarking anatomical accuracy, compositional reasoning, and trustworthiness in generative models.

Recent text-to-image diffusion models have demonstrated remarkable generation of realistic facial images conditioned on textual prompts and human identities, enabling creating personalized facial imagery. However, existing prompt-based methods for removing or modifying identity-specific features rely either on the subject being well-represented in the pre-trained model or require model fine-tuning for specific identities. In this work, we analyze the identity generation process and introduce a reverse personalization framework for face anonymization. Our approach leverages conditional diffusion inversion, allowing direct manipulation of images without using text prompts. To generalize beyond subjects in the model's training data, we incorporate an identity-guided conditioning branch. Unlike prior anonymization methods, which lack control over facial attributes, our framework supports attribute-controllable anonymization. We demonstrate that our method achieves a state-of-the-art balance between identity removal, attribute preservation, and image quality. Source code and data are available at https://github.com/hanweikung/reverse-personalization

We propose Cross-Attention-based Non-local Knowledge Distillation (CanKD), a novel feature-based knowledge distillation framework that leverages cross-attention mechanisms to enhance the knowledge transfer process. Unlike traditional self-attention-based distillation methods that align teacher and student feature maps independently, CanKD enables each pixel in the student feature map to dynamically consider all pixels in the teacher feature map. This non-local knowledge transfer more thoroughly captures pixel-wise relationships, improving feature representation learning. Our method introduces only an additional loss function to achieve superior performance compared with existing attention-guided distillation methods. Extensive experiments on object detection and image segmentation tasks demonstrate that CanKD outperforms state-of-the-art feature and hybrid distillation methods. These experimental results highlight CanKD's potential as a new paradigm for attention-guided distillation in computer vision tasks. Code is available at https://github.com/tori-hotaru/CanKD

Image generation models frequently encode social biases, including stereotypes tied to gender, race, and profession. Existing methods for analyzing these biases in diffusion models either focus narrowly on predefined categories or depend on manual interpretation of latent directions. These constraints limit scalability and hinder the discovery of subtle or unanticipated patterns. We introduce SCALEX, a framework for scalable and automated exploration of diffusion model latent spaces. SCALEX extracts semantically meaningful directions from H-space using only natural language prompts, enabling zero-shot interpretation without retraining or labelling. This allows systematic comparison across arbitrary concepts and large-scale discovery of internal model associations. We show that SCALEX detects gender bias in profession prompts, ranks semantic alignment across identity descriptors, and reveals clustered conceptual structure without supervision. By linking prompts to latent directions directly, SCALEX makes bias analysis in diffusion models more scalable, interpretable, and extensible than prior approaches.

Innovations in digital intelligence are transforming robotic surgery through more informed decision-making. Real-time awareness of surgical instrument presence and actions (e.g., cutting tissue) is essential, yet despite decades of research, most machine learning models rely on small datasets and still struggle to generalize. Recently, Vision-Language Models (VLMs) have achieved transformative advances in multimodal reasoning, suggesting strong potential for intelligent robotic surgery. However, surgical VLMs remain underexplored, and existing models show limited performance, underscoring the need for systematic benchmarks to assess their capabilities, limitations, and future development. To this end, we benchmark the zero-shot performance of several advanced VLMs on two public robotic-assisted laparoscopic datasets for instrument and action classification. Beyond standard evaluation, we integrate explainable AI to visualize VLM attention and uncover causal explanations behind predictions, providing a previously underexplored perspective for assessing model reliability. We also propose explainability-based metrics to complement standard evaluations. Our analysis reveals that surgical VLMs, despite domain-specific training, often rely on weak contextual cues rather than clinically meaningful visual evidence, highlighting the need for stronger visual and reasoning supervision in surgical applications. The code is provided in our public repository at https://github.com/jiajun344/SurgXBench-Explainable-Vision-Language-Model-Benchmark-for-Surgery

While recent text-to-image models excel at generating realistic content, they struggle to capture the nuanced visual characteristics that define a brand's distinctive style--such as lighting preferences, photography genres, color palettes, and compositional choices. This work introduces BrandFusion, a novel framework that automatically generates brand-aligned promotional images by decoupling brand style learning from image generation. Our approach consists of two components: a Brand-aware Vision-Language Model (BrandVLM) that predicts brand-relevant style characteristics and corresponding visual embeddings from marketer-provided contextual information, and a Brand-aware Diffusion Model (BrandDM) that generates images conditioned on these learned style representations. Unlike existing personalization methods that require separate fine-tuning for each brand, BrandFusion maintains scalability while preserving interpretability through textual style characteristics. Our method generalizes effectively to unseen brands by leveraging common industry sector-level visual patterns. Extensive evaluation demonstrates consistent improvements over existing approaches across multiple brand alignment metrics, with a 68.61% preference rate in human evaluation study. This work paves the way for AI-assisted on-brand content creation in marketing workflows.

Generating synthetic multi-view images from a text prompt is an essential bridge to generating synthetic 3D assets. In this work, we introduce RapidMV, a novel text-to-multi-view generative model that can produce 32 multi-view synthetic images in just around 5 seconds. In essence, we introduce a novel spatio-angular latent space, where we encode not only the spatial appearance of a single frame, but also the angular viewpoint deviations across multiple frames into a single latent for improved efficiency and multi-view consistency. We achieve effective training of RapidMV by strategically decomposing our training process into multiple steps. We demonstrate that RapidMV outperforms existing methods in terms of consistency and latency, with competitive quality and text-image alignment.

Realistic and smooth full-body tracking is crucial for immersive AR/VR applications. Existing systems primarily track head and hands via Head Mounted Devices (HMDs) and controllers, making the 3D full-body reconstruction incomplete. One potential approach is to generate the full-body motions from sparse inputs collected from limited sensors using a Neural Network (NN) model. In this paper, we propose a novel method based on a multi-layer perceptron (MLP) backbone that is enhanced with residual connections and a novel NN-component called Memory-Block. In particular, Memory-Block represents missing sensor data with trainable code-vectors, which are combined with the sparse signals from previous time instances to improve the temporal consistency. Furthermore, we formulate our solution as a multi-task learning problem, allowing our MLP-backbone to learn robust representations that boost accuracy. Our experiments show that our method outperforms state-of-the-art baselines by substantially reducing prediction errors. Moreover, it achieves 72 FPS on mobile HMDs that ultimately improves the accuracy-running time tradeoff.

3D Gaussian Splatting (GS) is one of the most promising novel 3D representations that has received great interest in computer graphics and computer vision. While various systems have introduced editing capabilities for 3D GS, such as those guided by text prompts, fine-grained control over deformation remains an open challenge. In this work, we present a novel sketch-guided 3D GS deformation system that allows users to intuitively modify the geometry of a 3D GS model by drawing a silhouette sketch from a single viewpoint. Our approach introduces a new deformation method that combines cage-based deformations with a variant of Neural Jacobian Fields, enabling precise, fine-grained control. Additionally, it leverages large-scale 2D diffusion priors and ControlNet to ensure the generated deformations are semantically plausible. Through a series of experiments, we demonstrate the effectiveness of our method and showcase its ability to animate static 3D GS models as one of its applications.

Autonomous driving requires understanding infrastructure elements, such as lanes and crosswalks. To navigate safely, this understanding must be derived from sensor data in real-time and needs to be represented in vectorized form. Learned Bird's-Eye View (BEV) encoders are commonly used to combine a set of camera images from multiple views into one joint latent BEV grid. Traditionally, from this latent space, an intermediate raster map is predicted, providing dense spatial supervision but requiring post-processing into the desired vectorized form. More recent models directly derive infrastructure elements as polylines using vectorized map decoders, providing instance-level information. Our approach, Augmentation Map Network (AugMapNet), proposes latent BEV feature grid augmentation, a novel technique that significantly enhances the latent BEV representation. AugMapNet combines vector decoding and dense spatial supervision more effectively than existing architectures while remaining easy to integrate compared to other hybrid approaches. It additionally benefits from extra processing on its latent BEV features. Experiments on nuScenes and Argoverse2 datasets demonstrate significant improvements on vectorized map prediction of up to 13.3% over the StreamMapNet baseline on 60 m range and greater improvements on larger ranges. We confirm transferability by applying our method to another baseline, SQD-MapNet, and find similar improvements. A detailed analysis of the latent BEV grid confirms a more structured latent space of AugMapNet and shows the value of our novel concept beyond pure performance improvement. The code can be found at https://github.com/tmonnin/augmapnet

Exemplar-free class-incremental learning (EFCIL) aims to retain old knowledge acquired in the previous task while learning new classes, without storing the previous images due to storage constraints or privacy concerns. In EFCIL, the plasticity-stability dilemma, learning new tasks versus catastrophic forgetting, is a significant challenge, primarily due to the unavailability of images from earlier tasks. In this paper, we introduce adversarial pseudo-replay (APR), a method that perturbs the images of the new task with adversarial attack, to synthesize the pseudo-replay images online without storing any replay samples. During the new task training, the adversarial attack is conducted on the new task images with augmented old class mean prototypes as targets, and the resulting images are used for knowledge distillation to prevent semantic drift. Moreover, we calibrate the covariance matrices to compensate for the semantic drift after each task, by learning a transfer matrix on the pseudo-replay samples. Our method reconciles stability and plasticity, achieving state-of-the-art on challenging cold-start settings of the standard EFCIL benchmarks.

Contrastive learning methods in computer vision typically rely on augmented views of the same image or multimodal pretraining strategies that align paired modalities. However, these approaches often overlook semantic relationships between distinct instances, leading to false negatives when semantically similar samples are treated as negatives. This limitation is especially critical in medical imaging domains such as cardiology, where demographic and clinical attributes play a critical role in assessing disease risk and patient outcomes. We introduce a tabular-guided contrastive learning framework that leverages clinically relevant tabular data to identify patient-level similarities and construct more meaningful pairs, enabling semantically aligned representation learning without requiring joint embeddings across modalities. Additionally, we adapt the k-NN algorithm for zero-shot prediction to overcome the lack of zero-shot capability in unimodal representations. We demonstrate the strength of our methods using a large cohort of short-axis cardiac MR images and clinical attributes, where tabular data helps to more effectively distinguish between patient subgroups. Evaluation on downstream tasks, including fine-tuning, linear probing, and zero-shot prediction of cardiovascular artery diseases and cardiac phenotypes, shows that incorporating tabular data guidance yields stronger visual representations than conventional methods that rely solely on image augmentation or combined image-tabular embeddings. Further, we show that our method can generalize to natural images by evaluating it on a car advertisement dataset. Code is available at https://github.com/marteczkah/tables_guide_vision.

Chart understanding is crucial for deploying multimodal large language models (MLLMs) in real-world scenarios such as analyzing scientific papers and technical reports. Unlike natural images, charts pair a structured visual layout (spatial property) with an underlying data representation (textual property) -- grasping both is essential for precise, fine-grained chart reasoning.Motivated by this observation, we propose START, the Spatial and Textual learning for chART understanding. Specifically, we introduce (i) chart-element grounding and (ii) chart-to-code generation to strengthen an MLLM's understanding of both chart visual layout and data details. To facilitate spatial and textual learning, we propose the START-Dataset generated with a novel data-generation pipeline that first leverages a vision-language model (VLM) to translate real chart images into executable chart code, recovering the underlying data representation while preserving the visual distribution of real-world charts. We then evolve the code with a Large Language Model (LLM) to ascertain the positions of chart elements that capture the chart's visual structure, addressing challenges that existing methods cannot handle. To evaluate a model's ability to understand chart spatial structures, we propose the Chart Spatial understanding Benchmark (CS-Bench), filling a critical gap in comprehensive chart understanding evaluation.Leveraging spatial and textual learning, START delivers consistent gains across model sizes and benchmarks over the base models and surpasses prior state-of-the-art by a clear margin. Code, data and models will be publicly available.

We propose two algorithms for 3D symmetry detection based on enhanced back-projection of vision features extracted from foundation vision models such as DINOv2. Our method enhances back-projection by rendering multiple views of 3D objects, extracting features, and projecting them onto the geometry with two key improvements--Fibonacci view sampling and view rotations--that increase robustness and accuracy. Using these features, we detect symmetry planes and axes through two dedicated algorithms. Experiments on ShapeNet show that our plane detection approach outperforms both traditional geometric and learning-based methods by a wide margin. The method is also efficient, running in seconds on a single 8GB GPU, making it practical for large-scale or real-world applications. Overall, our results demonstrate that enhanced back-projection of vision features offers a simple yet effective framework for solving fundamental 3D geometric problems such as symmetry detection. Code is available at https://github.com/Spulp/EnhancedBackProjection.

Deep learning models often struggle to maintain performance when the training and testing data come from different distributions. Test-time adaptation (TTA) addresses this by adapting a pre-trained model to an unlabeled target domain under distribution shifts. A more challenging setting is open-set TTA (OSTTA), where the target domain may contain unknown samples outside the source classes. Existing OSTTA methods primarily detect and discard such unknowns, relying only on known samples for adaptation. In this work, we argue that unknown samples can also provide valuable cues for improving adaptation. We propose LU-OSTTA (learning from unknown for OSTTA), a simple yet effective framework that leverages both in-distribution and semantically useful out-of-distribution samples. Our approach introduces: (i) a class-conditioned dynamic energy threshold to separate OOD samples more reliably, (ii) an optimal transport-based pseudo-label refinement to mitigate noise under distribution shifts, and (iii) an adaptive prototype weighting strategy that emphasizes semantically aligned target samples while down-weighting harmful ones. Experimental results demonstrate that our LU-OSTTA consistently outperforms state-of-the-art TTA and OSTTA methods, highlighting the benefits of utilizing rather than discarding unknown samples. Our code is available at: https://github.com/takihasan/LU-OSTTA.

Accurate, up-to-date High-Definition (HD) maps are critical for urban planning, infrastructure monitoring, and autonomous navigation. However, these maps quickly become outdated as environments evolve, creating a need for robust methods that not only detect changes but also incorporate them into updated 3D representations. While change detection techniques have advanced significantly, there remains a clear gap between detecting changes and actually updating 3D maps, particularly when relying on 2D image-based change detection. To address this gap, we introduce SceneEdited, the first city-scale dataset explicitly designed to support research on HD map maintenance through 3D point cloud updating. SceneEdited contains over 800 up-to-date scenes covering 73 km of driving and approximate 3 \text km ^2 of urban area, with more than 23,000 synthesized object changes created both manually and automatically across 2000+ out-of-date versions, simulating realistic urban modifications such as missing roadside infrastructure, buildings, overpasses, and utility poles. Each scene includes calibrated RGB images, LiDAR scans, and detailed change masks for training and evaluation. We also provide baseline methods using a foundational image-based structure-from-motion pipeline for updating outdated scenes, as well as a comprehensive toolkit supporting scalability, trackability, and portability for future dataset expansion and unification of out-of-date object annotations. Both the dataset and the toolkit are publicly available at https://github.com/ChadLin9596/ScenePoint-ETK, establising a standardized benchmark for 3D map updating research.

Event cameras are increasingly being explored for space applications due to their high dynamic range and increased spatiotemporal resolution. Existing datasets in this application have focused on capturing low-light, sub-pixel space objects and Earth observation scenarios. There remains a notable gap in datasets tailored to high-illumination conditions, particularly those involving direct solar imaging. This work introduces a dataset of solar event recordings captured with an event camera in a controlled sun-simulator environment. The dataset is specifically designed to support research in sun sensing and stray light analysis for spacecraft attitude estimation applications. It includes raw event data, annotated sun centroid locations, object motion profiles, and secondary optical aberration artifacts. In addition to the dataset, we present a systematic methodology for estimating the sun vector, intended to serve as a benchmark for evaluating sun sensing approaches in this application. All data and code are open source to facilitate further study.

In current research, Bird's-eye-view (BEV)-based transformers are increasingly utilized for multi-camera 3D object detection. Traditional models often employ random queries as anchors, optimizing them successively. Recent advancements complement or replace these random queries with detections from auxiliary networks. We propose a more intuitive and efficient approach by using BEV feature cells directly as anchors. This end-to-end approach leverages the dense grid of BEV queries, considering each cell as a potential object for the final detection task. As a result, we introduce a novel two-stage anchor generation method specifically designed for multi-camera 3D object detection. To address the scaling issues of attention with a large number of queries, we apply BEV-based Non-Maximum Suppression, allowing gradients to flow only through non-suppressed objects. This ensures efficient training without the need for post-processing. By using BEV features from encoders such as BEVFormer directly as object queries, temporal BEV information is inherently embedded. Building on the temporal BEV information already embedded in our object queries, we introduce a hybrid temporal modeling approach by integrating prior detections to further enhance detection performance. Evaluating our method on the nuScenes dataset shows consistent and significant improvements in NDS and mAP over the baseline, even with sparser BEV grids and therefore fewer initial anchors. It is particularly effective for small objects, enhancing pedestrian detection with a 3.8% mAP increase on nuScenes and an 8% increase in LET-mAP on Waymo. Applying our method, named DenseBEV, to the challenging Waymo Open dataset yields state-of-the-art performance, achieving a LET-mAP of 60.7%, surpassing the previous best by 5.4%. Code is available at https://github.com/mdaehl/DenseBEV.