Sparse-view 3D reconstruction is increasingly addressed with feed-forward splatting networks that predict explicit primitives directly from images. Yet most existing methods remain centered on Gaussian primitives and expose surfaces only indirectly: extracting a usable mesh for downstream simulation, physics reasoning, or embodied interaction still requires expensive post-hoc steps that break the feed-forward promise. This limitation is especially pronounced in pose-free settings, where scene structure and camera parameters must be estimated jointly from sparse observations. We present TriSplat, a feed-forward reconstruction network that represents scenes with oriented triangle primitives and directly exports simulation-ready mesh scenes from a single forward pass. Given input images, the network predicts local 3D point maps, triangle attributes, camera poses, and optional intrinsics. Rather than regressing triangle orientation as an unconstrained latent variable, our approach constructs geometry normals from the predicted point maps, refines them with an image-conditioned normal head, and converts them into stable local frames for triangle parameterization. A mono-normal bootstrap schedule further stabilizes early training, while opacity and blur scheduling progressively sharpens the learned surface representation for direct mesh extraction. Experiments on RealEstate10K and DL3DV show that this representation produces more geometry-faithful reconstructions than Gaussian feed-forward baselines while maintaining competitive novel-view rendering quality. Because the rendering primitives are themselves surface triangles, the output can be directly ingested by physics engines, collision detectors, and standard rendering pipelines without any conversion, making it a practical simulation-ready solution for feed-forward 3D scene reconstruction.

Autonomous robots often need to move their camera before they can act: to inspect an object, reveal an occluded region, or obtain a view that responds to a user's intent. While vision-language navigation translates instructions to base motion and vision-language-action policies map instructions to manipulation actions, language-conditioned camera motion remains comparatively underexplored as a first-class action. We formulate language-conditioned camera motion generation: given a current RGB observation and a free-form natural-language intent, predict a relative target camera pose for the next observation. This task is inherently non-trivial: viewpoint changes are driven by latent perceptual intentions, and a valid motion may operate at different semantic granularity, from entering a room to looking around a corner, inspecting a visible object, or revealing an occluded detail. To model this structure, we mine multi-intention camera-motion supervision from egocentric video, pairing plausible intents and observation-gain descriptions with relative SE(3) target poses. We propose LIME, a vision-language camera-motion generator that combines an auto-regressive observation-gain output with a continuous flow-matching pose head. This design lets the model jointly predict what the next view should reveal while representing multi-hypothesis target views. Across experiments and downstream robotic tasks, we show that LIME can learn to actively choose camera poses from passive human video, turning ordinary egocentric recordings into supervision for intent-aware active perception.

We present DynHair, a novel method for tracking and modeling dynamic hair for human head avatars. From video input, we reconstruct a dynamic head avatar with an explicit strand-based hair representation using structured 3D Gaussian Splatting. In contrast to the face region of human head avatars, which can be modeled with 3D Gaussians that are attached or generated with respect to some expressive 3D head model, hair is particularly challenging as it exhibits dynamic motion effects. Therefore, we present a novel method that models the dynamic deformations of the hair strands using a temporal network that is conditioned on angular velocity and acceleration of the head, as well as relative gravity. Specifically, an LSTM encodes the motion history and modulates per-point strand features via FiLM conditioning which further used by MLP to produce physically plausible displacements to canonical hairstyle. We jointly optimize this motion and appearance representation of the hair, with a 3DGS-based representation of the face-region, via differentiable Gaussian splatting with photometric, geometric, and physics-based supervision. As a result of our method, we retrieve hair tracking of the training video data and an animatable head avatar with controllable hair dynamics. In our experiments, we demonstrate state-of-the-art performance in terms of hair dynamics, temporal consistency, and generalization across subjects.

Visual odometry (VO) is fundamental to autonomous navigation, robotics, and augmented reality. While self-supervised learning has eliminated the need for expensive ground-truth labels in monocular VO, dynamic objects and occlusions that violate the static scene assumption lead to erroneous pose estimates. Existing uncertainty-based methods filter unreliable regions but rely solely on single-frame information, neglecting temporal consistency across consecutive frames. We present Combined Projected Uncertainty (CoProU), a principled probabilistic formulation that propagates and fuses uncertainties across temporal frames. Our key insight is that robust uncertainty estimation requires combining target frame uncertainty with projected uncertainty from reference frames, enabling effective identification of dynamic regions and temporal inconsistencies. We demonstrate CoProU’s versatility through two complementary frameworks. CoProU-VO-2F employs a decoupled architecture with CNN-based pose encoder and vision transformer-based depth encoder for two-frame visual odometry. CoProU-VO-MF extends our approach to multi-frame scenarios using a unified transformer architecture with coupled encoders that produce shared representations for ego-motion and geometry estimation. This demonstrates that CoProU, though originally formulated for frame pairs, generalizes naturally to multi-frame settings through pairwise application. Comprehensive experiments validate our contributions. CoProU-VO-2F achieves substantial improvements over state-of-the-art two-frame methods, reducing ATE by up to 63 Project Page .

Target localization is a prerequisite for embodied tasks such as navigation and manipulation. Conventional approaches rely on constructing explicit 3D scene representations to enable target localization, such as point clouds, voxel grids, or scene graphs. While effective, these pipelines incur substantial mapping time, storage overhead, and scalability limitations. Recent advances in vision-language models suggest that rich semantic reasoning can be performed directly on 2D observations, raising a fundamental question: is a complete 3D scene reconstruction necessary for object localization? In this work, we revisit object localization and propose a map-free pipeline that stores only posed RGB-D keyframes as a lightweight visual memory–without constructing any global 3D representation of the scene. At query time, our method retrieves candidate views, re-ranks them with a vision-language model, and constructs a sparse, on-demand 3D estimate of the queried target through depth backprojection and multi-view fusion. Compared to reconstruction-based pipelines, this design drastically reduces preprocessing cost, enabling scene indexing that is over two orders of magnitude faster to build while using substantially less storage. We further validate the localized targets on downstream object-goal navigation tasks. Despite requiring no task-specific training, our approach achieves strong performance across multiple benchmarks, demonstrating that direct reasoning over image-based scene memory can effectively replace dense 3D reconstruction for object-centric robot navigation. Project page: https://ruizhou-cn.github.io/memory-over-maps/

Despite rapid progress, pretrained vision-language models still struggle when answers depend on tiny visual details or on combining clues spread across multiple regions, as in documents and compositional queries. We address this by framing grounding as test-time evidence retrieval: given a query, the model should actively identify where to look next to resolve ambiguity. To this end, we propose a training-free, model-intrinsic grounding method that uses uncertainty as supervision. Specifically, we compute the entropy of the model's next-token distribution and backpropagate it to the visual token embeddings to obtain an entropy-gradient relevance map, without auxiliary detectors or attention-map heuristics. We then extract and rank multiple coherent regions to support multi-evidence queries, and introduce an iterative zoom-and-reground procedure with a spatial-entropy stopping rule to avoid over-refinement. Experiments on seven benchmarks across four VLM architectures demonstrate consistent improvements over existing methods, with the largest gains on detail-critical and high-resolution settings, while also producing more interpretable evidence localizations.

Existing visual localization methods are typically either 2D image-based, which are easy to build and maintain but limited in effective geometric reasoning, or 3D structure-based, which achieve high accuracy but require a centralized reconstruction and are difficult to update. In this work, we revisit visual localization with a 2D image-based representation and propose to augment each image with estimated depth maps to capture the geometric structure. Supported by the effective use of dense matchers, this representation is not only easy to build and maintain, but achieves highest accuracy in challenging conditions. With compact compression and a GPU-accelerated LO-RANSAC implementation, the whole pipeline is efficient in both storage and computation and allows for a flexible trade-off between accuracy and highest memory efficiency. Our method achieves a new state-of-the-art accuracy on various standard benchmarks and outperforms existing memory-efficient methods at comparable map sizes. Code will be available at https://github.com/cvg/Hierarchical-Localization.

Accurate and automated data analysis for transmission electron microscopy will enable new high-throughput experiments that can reveal atomic-scale structure-property relationships for many functional materials. A key challenge in this pursuit is scalable three-dimensional structure prediction from single two-dimensional images. Existing tomographic and atom-counting approaches require either high electron doses, complex acquisition schemes, or the object in specific orientations, limiting experimental design. Here, we introduce a diffusion-based generative workflow that predicts the three-dimensional morphology of nanoscale objects directly from a single scanning/transmission electron micrograph. Applied to sub-5 nm platinum nanoparticles on ceria, it successfully predicts reasonable structures across diverse particle morphologies and imaging orientations. Combined with automated data acquisition in operando experiments, we believe techniques like this could be an essential part in relating ensemble-level structural variation and dynamics with performance, particularly fitting for heterogeneous catalysis.

Autonomous inspection is a central problem in robotics, with applications ranging from industrial monitoring to search-and-rescue. Traditionally, inspection has often been reduced to navigation tasks, where the objective is to reach a predefined location while avoiding obstacles. However, this formulation captures only part of the real inspection problem. In real-world environments, the inspection targets may become visible well before their exact coordinates are reached, making further movement both redundant and inefficient. What matters more for inspection is not simply arriving at the target’s position, but positioning the robot at a viewpoint from which the target becomes observable. In this work, we revisit inspection from a perception-aware perspective. We propose an end-to-end reinforcement learning framework that explicitly incorporates target visibility as the primary objective, enabling the robot to find the shortest trajectory that guarantees visual contact with the target without relying on a map. The learned policy leverages both perceptual and proprioceptive sensing and is trained entirely in simulation, before being deployed to a real-world robot. We further develop an algorithm to compute ground-truth shortest inspection paths, which provides a reference for evaluation. Through extensive experiments, we show that our method outperforms existing classical and learning-based navigation approaches, yielding more efficient inspection trajectories in both simulated and real-world settings.

This paper introduces RaCo, a lightweight neural network designed to learn robust and versatile keypoints suitable for a variety of 3D computer vision tasks. The model integrates three key components: the repeatable keypoint detector, a differentiable ranker to maximize matches with a limited number of keypoints, and a covariance estimator to quantify spatial uncertainty in metric scale. Trained on perspective image crops only, RaCo operates without the need for covisible image pairs. It achieves strong rotational robustness through extensive data augmentation, even without the use of computationally expensive equivariant network architectures. The method is evaluated on several challenging datasets, where it demonstrates state-of-the-art performance in keypoint repeatability and two-view matching, particularly under large in-plane rotations. Ultimately, RaCo provides an effective and simple strategy to independently estimate keypoint ranking and metric covariance without additional labels, detecting interpretable and repeatable interest points. The code is available at https://github.com/cvg/RaCo.

We propose Unblur-SLAM, an RGB SLAM pipeline for sharp 3D reconstruction from blurred image inputs. In contrast to previous work, our approach is able to handle different types of blur and demonstrates state-of-the-art performance in the presence of both motion blur and defocus blur. Moreover, we adjust the computation effort with the amount of blur in the input image.As a first stage, our method uses a feed-forward image deblurring model for which we propose a suitable training scheme that can improve both tracking and mapping modules.Frames that are successfully deblurred by the feed-forward network obtain refined poses and depth through local-global multi-view optimization and loop closure. Frames that fail the first stage deblurring are directly modeled through the global 3DGS representation and an additional blur network to model multiple blurred sub-frames and simulate the blur formation process in 3D space, thereby learning sharp details and refined sub-frame poses.Experiments on several real-world datasets demonstrate consistent improvements in both pose estimation and sharp reconstruction results of geometry and texture.

Video Language Models (VideoLMs) enable AI systems to understand temporal dynamics in videos. To fit within the maximum context window constraint, current methods use keyframe sampling which often misses both macro-level events and micro-level details due to the sparse temporal coverage. Furthermore, processing full images and their tokens for each frame incurs substantial computational overhead. We address these limitations by leveraging video codec primitives (specifically motion vectors and residuals) which natively encode video redundancy and sparsity without requiring expensive full-image encoding for most frames. To this end, we introduce lightweight transformer-based encoders that aggregate codec primitives and align their representations with image encoder embeddings through a pre-training strategy that accelerates convergence during end-to-end fine-tuning. Our approach, CoPE-VideoLM, reduces the time-to-first-token by up to 86

Real-time monocular 3D object detection remains challenging due to severe depth ambiguity, viewpoint shifts, and the high computational cost of 3D reasoning. Existing approaches either rely on LiDAR or geometric priors to compensate for missing depth or sacrifice efficiency to achieve competitive accuracy. We introduce LeAD-M3D, a monocular 3D detector that achieves state-of-the-art accuracy and real-time inference without extra modalities. Our method is enabled by three key components. Asymmetric Augmentation Denoising Distillation (A2D2) transfers geometric knowledge from a clean-image teacher to a MixUp-noised student via a quality- and importance-weighted depth-feature loss, enabling stronger depth reasoning without LiDAR. 3D-aware Consistent Matching (CM_3D) improves prediction-to-ground truth assignment by integrating 3D MGIoU into the matching score, yielding stable and precise supervision. Finally, Confidence-Gated 3D Inference (CGI_3D) accelerates inference by restricting expensive 3D regression to confident regions. Together, these contributions set a new Pareto frontier for monocular 3D detection: LeAD-M3D achieves state-of-the-art accuracy on KITTI and Waymo, and the best reported car AP on Rope3D, while running up to 3.6× faster than prior high-accuracy models (e.g., MonoDiff). LeAD-M3D demonstrates that high fidelity and real-time monocular 3D detection is simultaneously attainable, without LiDAR, stereo, or strong geometric assumptions.

Urban perception describes how people subjectively evaluate urban environments, shaping how cities are experienced and understood. Existing computational approaches primarily model urban perception directly from street view images, but largely ignore the human perceptual process through which such judgments are formed. In this paper, we introduce Place Pulse-Gaze, an urban perception dataset that augments street view images with synchronized eye-tracking recordings and individual perception labels. Based on this dataset, we propose a Gaze-Guided Urban Perception Framework to study how gaze behavior contributes to the modeling of subjective urban perception. The framework systematically investigates three complementary settings: gaze-only modeling, gaze fusion with explicit semantic scene representations, and gaze fusion with implicit richer visual representations. Experiments show that gaze alone already carries useful predictive signals for subjective urban perception, and that integrating gaze with scene representations further improves prediction under both semantic and richer visual representations. Overall, our findings highlight the importance of incorporating human perceptual processes into urban scene understanding and open a direction for gaze-guided multimodal urban computing.

Accurate spatiotemporal alignment of multi-view video streams is essential for a wide range of dynamic-scene applications such as multi-view 3D reconstruction, pose estimation, and scene understanding. However, synchronizing multiple cameras remains a significant challenge, especially in heterogeneous setups combining professional and consumer-grade devices, visible and infrared sensors, or systems with and without audio, where common hardware synchronization capabilities are often unavailable. This limitation is particularly evident in real-world environments, where controlled capture conditions are not feasible. In this work, we present a low-cost, general-purpose synchronization method that achieves millisecond-level temporal alignment across diverse camera systems while supporting both visible (RGB) and infrared (IR) modalities. The proposed solution employs a custom-built LED Clock that encodes time through red and infrared LEDs, allowing visual decoding of the exposure window (start and end times) from recorded frames for millisecond-level synchronization. We benchmark our method against hardware synchronization and achieve a residual error of 1.34 ms RMSE across multiple recordings. In further experiments, our method outperforms light-, audio-, and timecode-based synchronization approaches and directly improves downstream computer vision tasks, including multi-view pose estimation and 3D reconstruction. Finally, we validate the system in large-scale surgical recordings involving over 25 heterogeneous cameras spanning both IR and RGB modalities. This solution simplifies and streamlines the synchronization pipeline and expands access to advanced vision-based sensing in unconstrained environments, including industrial and clinical applications.

Human-human collaboration is a fundamental aspect of everyday life, essential to success in a wide range of goal-directed activities from household tasks to professional teamwork. While much research has focused on modeling coordination and task execution, the cognitive processes that support such collaboration, particularly Theory of Mind (the ability to infer the mental states of others), remain difficult to study in natural settings. To address this gap, we introduce a novel egocentric and exocentric video dataset capturing real-world collaboration in cooking scenarios. The dataset integrates multi-perspective video, high-quality audio, gaze tracking, and 3D scene and object scans, with annotations for shared attention to objects, social cues and interactions between agents, as well as agent-object interactions. We establish benchmarks for Joint Attention Estimation, Socially Conditioned Object Interaction Anticipation, and Collaborative Handover Prediction, enabling research on multimodal perception, proactive assistance, and collaborative planning. By providing temporally aligned, richly annotated multimodal data, CoMind facilitates the development and evaluation of AI systems capable of modeling complex social interactions and reasoning about human behaviors in collaborative environments. Our dataset and benchmarks are made available at https://comind.ethz.ch/.

We propose SNI-SLAM++, a tightly-coupled semantic SLAM system utilizing neural implicit representation, that simultaneously performs accurate semantic mapping, high-quality surface reconstruction, and robust camera tracking. Our system tightly integrates visual appearance, geometry, and semantics through five key components: (i) We introduce hierarchical semantic representation to allow multi-level semantic comprehension for top-down structured semantic mapping of the scene. (ii) To fully utilize the correlation between multiple attributes of the environment, we integrate appearance, geometry and semantic features through cross-attention for feature collaboration. This strategy enables a more multifaceted understanding of the environment, thereby allowing SNI-SLAM++ to remain robust even when single attribute is defective. (iii) We design an internal fusion-based decoder to obtain semantic, RGB, and Truncated Signed Distance Field (TSDF) values from multi-level features for accurate decoding. (iv) We introduce a semantics-coupled tracking framework that tightly incorporates semantic constraints for camera pose estimation in neural implicit SLAM. This framework leverages the multi-view consistency of semantics to construct a pose graph and perform semantic loop closure optimization, enabling robust tracking. (v) We propose a feature loss to update the scene representation at the feature level. Compared with low-level losses such as RGB loss and depth loss, our feature loss is capable of guiding the network optimization on a higher level. Our SNI-SLAM++ demonstrates superior performance over all recent visual SLAM methods in terms of mapping and tracking accuracy on the datasets of Replica, ScanNet, TUM-RGBD, and ScanNet++, while also showing excellent capabilities in accurate semantic segmentation and 3D semantic mapping.

Generative novel view synthesis from sparse input images is rarely all reconstruction or all generation: pixels visible in some source view have a unique correct value modulated only by view-dependent shading, while pixels in disocclusions or beyond the captured volume admit a distribution of plausible completions. Existing generative novel-view-synthesis methods conflate these regimes under a single uniform loss, blurring the line between geometric fidelity and creative hallucinations even when scene geometry is injected through warped point clouds or projected depth. We introduce GenRec, a multi-view flow matching model that builds the reconstruction–generation split directly into its architecture, supervision, and gradient flow. Guided by an observation mask derived from the source cameras and a monocular depth estimator, a flow matching backbone jointly denoises RGB and scene-coordinate maps across all target views, while a pixel-space refinement stage restores high-frequency detail on observed pixels; the same mask gates supervision so regression signals do not contaminate the generative prior. Across RealEstate10K, DL3DV-10K, and Mip-NeRF 360, in both single-view extrapolation and two-view interpolation, GenRec attains the best reconstruction fidelity in observed regions while also surpassing purely generative baselines on perceptual quality in unobserved ones, showing the effectiveness of our approach.

We present a dataset for force-grounded, cross-view articulated manipulation that couples what is seen with what is done and what is felt during real human interaction. The dataset contains 3048 sequences across 381 articulated objects in 38 environments. Each object is operated under four embodiments - (i) human hand, (ii) human hand with a wrist-mounted camera, (iii) handheld UMI gripper, and (iv) a custom Hoi! gripper - where the tool embodiment provide synchronized end-effector forces and tactile sensing. Our dataset offers a holistic view of interaction understanding from video, enabling researchers to evaluate how well methods transfer between human and robotic viewpoints, but also investigate underexplored modalities such as force sensing and prediction.