#01Nov 4, 2026
cs.CV
Expanding the Quality-Coverage Frontier with Competitive Flow Matching
Wenxuan Peng, Richard Zhang, Haitian Zheng and 6 more
Flow matching (FM) generators often face a quality-coverage tradeoff. Inference-time techniques such as classifier-free guidance raise quality, but at the cost of coverage. We hypothesize this limitation stems from a basic property of the objective: an intermediate noisy image could correspond to many clean images, yet the model can only predict a single velocity, at best averaging over the multiple possibilities. To mitigate this, we introduce competition among predictions during training. We achieve this by augmenting the predictor with a randomly drawn low-dimensional conditioning code $z$. For each training example, we sample $K$ values of $z$ to produce $K$ candidate velocities that compete to explain the target, and apply the loss only to the best-matching candidate. This winner-take-all objective allows different modes to shape different predictions, rather than being averaged into a single compromise. Furthermore, this term naturally enables a new form of guidance, auxiliary conditioning scaling, which complements traditional guidance methods. Together, these two guidance mechanisms show that our method substantially improves the quality-coverage frontier. We validate on ImageNet-256, using two different testbeds, SiT and RAEv2. Given less than 3% of the original SiT pretraining compute for post-training, at the 67% Recall point, our method reduces $\mathrm{FD}_{r,6}$ from 19.22 to 5.91. On RAEv2, at the 70% Recall point, it reduces $\mathrm{FD}_{r,6}$ from 16.83 to 3.18.
#02Nov 4, 2026
cs.CV
VINO: Revisiting 3D Representation Learning with Volume Transformers
Adrian Kruse, Kadir Yilmaz, Daan de Geus and 2 more
Self-supervised learning has given rise to vision foundation models that learn general-purpose image representations transferable across downstream tasks, yet representation learning for 3D scenes has not achieved comparable success. Motivated by this, we identify key limitations of current 3D representation learning approaches and revisit their core design choices. We first replace the commonly used hierarchical PTv3 backbone with the Volume Transformer, a ViT-like architecture for 3D scenes. We then redesign two core objectives of modern approaches built around student-teacher networks: matching representations across views and predicting masked content from visible context. For the first objective, we separate the roles of teacher and student: the teacher provides clean, consistent targets while the student learns invariance to stronger perturbations. For the second objective, we prevent the student from accessing the masked geometry and instead require it to infer missing representations from visible context only. Finally, we redesign 2D-to-3D knowledge distillation, replacing direct regression of view-dependent image features with view-consistent supervision. Together, these changes form VINO, a framework that learns semantic representations of 3D scenes without labels. VINO substantially improves dense linear probing across indoor and outdoor benchmarks and transfers strongly to out-of-domain settings.
#03Nov 4, 2026
cs.CV
From Pixels to Actions - Towards Native Vision-Language-Action Policies
Weichen Fan, Haiwen Diao, Penghao Wu and 14 more
Vision-language-action models are typically built on a simple premise: perception is first compressed by visual encoders, and action reasoning begins only afterwards. Yet manipulation often hinges on precisely the visual signals most vulnerable to this abstraction, e.g., local geometry, boundaries, depth, and subtle state changes. This raises a fundamental question: should action policies inherit a fixed visual endpoint, or directly access evolving visual computation? We introduce NEO-VLA, a structurally encoder-free vision-language-action framework. Images are tokenized by lightweight patch convolutional layers and evolve jointly with language throughout the backbone, allowing a flow-matching action expert to draw from instruction-conditioned visual representations at different depths. This simple change reveals a consistent pattern. Native architectures outperform encoder-based counterparts under controlled comparisons; full fine-tuning reduces but does not erase the gap, and additional scale alone cannot recover it. More strikingly, layer-wise probing shows that the action-critical signals often peak well before the final visual representation. Across diverse manipulation settings and distribution shifts, NEO-VLA consequently achieves stronger and more robust control. Together, these results suggest that embodied control benefits from access to visual representations across the computational hierarchy, rather than a fixed encoder output.
#04Nov 5, 2026
cs.CV
MIKU: Material Invariance and Knowledge Unification for PBR Generation
Yajie Bao, Jiaye Fu, Youtian Lin and 5 more
High-quality material generation for physically based rendering (PBR) remains challenging because datasets and synthesis pipelines lack standardized representations. Differences in material definitions, rendering conventions, and parameterizations lead to unstable and physically implausible results. We propose MIKU, a unified framework for standardized PBR asset representation and physically consistent material generation across diverse 3D content domains. Its core is a unified processing pipeline that standardizes the workflow from data preprocessing to final texture synthesis, maintaining consistent geometry and texture representations. A normalization scheme harmonizes heterogeneous material parameterizations in a common representation space, enabling robust cross-dataset training and inference. Building on this formulation, we train a novel physically aware diffusion model for PBR texture generation within a two-stage multi-view synthesis framework. First, geometry-aware albedo generation synthesizes multi-view albedo maps under explicit geometric guidance to ensure cross-view appearance consistency. Second, metallic-roughness generation uses the diffusion model to predict physically plausible roughness and metallic maps conditioned on the generated albedo and unified representation. Trained end-to-end on standardized data, the model learns to generate materials that respect category-level material priors and semantic part boundaries. By decoupling representation standardization from appearance synthesis, MIKU substantially improves material generation stability and physical consistency. Extensive experiments demonstrate state-of-the-art texture quality and semantic consistency across object parts, advancing automated material generation for 3D content creation.
#05Nov 4, 2026
cs.CV
Motion-guided Contrastive Regularization For Latent World Models
Yiwen Chen, Qiqi Hou, Jisoo Jeong and 3 more
JEPA-style world models learn action-conditioned latent dynamics without reconstructing observations, but distributional regularization alone does not ensure that their representations preserve information needed for control. We identify static collapse, a failure mode in which the encoder preserves episode-specific static content while suppressing manipulation-relevant state changes. Because static content is temporally predictable yet diverse across episodes, this failure can persist despite regularizers such as SIGReg. A controlled study on OGBench-Cube shows that introducing randomized backgrounds substantially degrades planning performance while leaving the underlying manipulation task unchanged. To address this problem, we propose motion-guided contrastive regularization. Our approach compares predicted latent transitions with transitions to candidate observations from the same episode, reducing the usefulness of static background identity for distinguishing action outcomes. Optical-flow similarities provide soft contrastive targets, allowing candidates with similar motion to receive similar supervision. Combined with latent prediction and SIGReg, our objective encourages representations to preserve within-episode state distinctions and similarities between related transitions. Our method improves planning success over LeWorldModel (LeWM) from 52% to 78% on OGBench-Cube with randomized backgrounds and from 14% to 54% on LIBERO-Object, while maintaining comparable performance on four other visual-control benchmarks.