Searcharxiv⌕ Search

arXiv subjects

Wenhua Jin

Publications and source records attributed to Wenhua Jin.

5 recordsLinked to original sources

Kosmo-GFM: A Production-Ready Geometry Foundation Model

We present Kosmo-GFM, a scene reconstruction system that integrates a geometry foundation model (GFM), a pose optimization engine (PoseEngine), and reconstruction-derived supervision from KosmoWorld. GFM predicts camera parameters, depth, cross-view correspondences, and dynamic 3D points from images and videos. PoseEngine combines these predictions with geometric optimization to recover scene-level camera poses and geometry for downstream 3D and 4D Gaussian Splatting. KosmoWorld generates images and geometric supervision from reconstructed real-world scenes to improve GFM's geometric predictions. On the Sparse and Medium subsets of SpatialBench, GFM achieves the lowest absolute relative depth error and the highest relative-pose AUC at 5 degrees among the compared methods. It also achieves the lowest mean endpoint errors for dynamic points and scene flow across the five evaluated datasets. Across 34 scenes completed by all compared reconstruction systems, PoseEngine achieves the highest scene-averaged pose AUC at thresholds of 1, 3, and 5 degrees, with a cumulative runtime of 14.07 hours, approximately 20% of GLUEMAP's 69.67 hours. PoseEngine Flash provides a further accuracy-efficiency trade-off, with a measured mixed-precision runtime of 2.62 hours, approximately 4% of GLUEMAP's runtime. Separate scaling experiments show generally favorable trends with increased data, model computation, and training budgets. These results support the use of learned geometric priors within a scene-level reconstruction pipeline.

cs.CV↗

Moving Regions, Not Moving Objects: Segmenting What Actually Moves under Camera Motion

Moving object segmentation is conventionally formulated at the sequence and object level, marking an entire instance as moving once it moves anywhere in a video. Recent instantaneous formulations distinguish moving from stationary intervals, but still generate masks at the object level through frozen segmentation foundation models. Because the mask is derived from an instance-level proposal, motion is typically attributed to the entire instance rather than localized to the regions that actually move. Moreover, such instance-level predictions can produce masks even when no part of the instance is actually moving, revealing a more fundamental issue: motion is not an intrinsic property of an instance; it can only be determined by comparing at least two frames. Building on this premise, we introduce MoRSe, which jointly interprets two frames to predict a moving-region mask, distinguishing actual scene motion from apparent motion induced by the camera. A single RGB encoder replaces multi-stage foundation stacks, so no optical flow, point tracks, or separate mask generator is needed at inference. We release MoRSe-5K, a benchmark of 42 sequences with Static, Dynamic-Static, and Dynamic settings designed to evaluate the model's ability to distinguish and localize moving and stationary regions under different camera and scene motion conditions. MoRSe outperforms both specialized moving object segmentation methods and dense 4D reconstruction methods with motion-thresholded predictions, while producing the fewest false positives on sequences where nothing moves.

cs.CV↗

UniQuery4R: Unified 4D Scene Reconstruction from a Single Query

Reconstructing dynamic 4D scenes requires jointly estimating correspondence, geometry, object motion, and camera motion. Existing feed-forward methods typically predict dense task-specific maps or independently process source-target pairs, leading to unnecessary computation for sparse queries and limited feature reuse across different frame pairs. We present UniQuery4R, a query-conditioned framework that encodes a multi-frame clip once and selects the source view, target view, and continuous source-image coordinate only at decoding time via source-to-target cross-attention. Each query jointly predicts target correspondence, target-time 3D position, and scene flow, along with source depth, while camera parameters are estimated per view. This design allows the encoded clip to be reused across arbitrary source-target selections and supports both sparse inference and dense reconstruction through batched queries, without learned temporal embeddings tied to a fixed clip length. We further introduce a direction-magnitude parameterization of scene flow with separate supervision for moving and static points. Among the evaluated methods, UniQuery4R achieves the best macro-average results on WorldTrack for both scene-flow estimation and dynamic-point reconstruction.

cs.CV↗

SpongeCake: A Layered Microflake Surface Appearance Model

In this paper, we propose SpongeCake: a layered BSDF model where each layer is a volumetric scattering medium, defined using microflake or other phase functions. We omit any reflecting and refracting interfaces between the layers. The first advantage of this formulation is that an exact and analytic solution for single scattering, regardless of the number of volumetric layers, can be derived. We propose to approximate multiple scattering by an additional single-scattering lobe with modified parameters and a Lambertian lobe. We use a parameter mapping neural network to find the parameters of the newly added lobes to closely approximate the multiple scattering effect. Despite the absence of layer interfaces, we demonstrate that many common material effects can be achieved with layers of SGGX microflake and other volumes with appropriate parameters. A normal mapping effect can also be achieved through mapping of microflake orientations, which avoids artifacts common in standard normal maps. Thanks to the analytical formulation, our model is very fast to evaluate and sample. Through various parameter settings, our model is able to handle many types of materials, like plastics, wood, cloth, etc., opening a number of practical applications.

cs.GR↗

Position-free Multiple-bounce Computations for Smith Microfacet BSDFs

Bidirectional Scattering Distribution Functions (BSDFs) encode how a material reflects or transmits the incoming light. The most commonly used model is the Microfacet BSDF. It computes material response from the micro-geometry of the surface assuming a single bounce on specular microfacets. The original model ignores multiple bounces on the micro-geometry, resulting in energy loss, especially with large roughness. In this paper, we present a position-free formulation of multiple bounces inside the micro-geometry, which eliminates this energy loss. We use an explicit mathematical definition of path space that describes single and multiple bounces in a uniform way. We then study the behavior of light on the different vertices and segments in path space, leading to an accurate and reciprocal multiple-bounce description of BSDFs. We also present practical, unbiased Monte-Carlo estimators to compute multiple scattering. Our method is less noisy than existing algorithms for computing multiple scattering. It is almost noise-free with a very-low sampling rate, from 2 to 4 samples per pixel.

cs.GR↗