Searcharxiv⌕ Search

arXiv subjects

Yunfan Zeng

Publications and source records attributed to Yunfan Zeng.

3 recordsLinked to original sources

CGOcc: Learning Cross-Gaussian Dependencies for 3D Semantic Occupancy Prediction

In Gaussian-based semantic occupancy prediction, overlapping primitives can provide ambiguous class predictions, making their joint interpretation important for voxel classification. However, current methods using spatially weighted mixtures neglect explicit modeling of cross-primitive dependencies, limiting the readout's ability to utilize category-specific relational evidence. Hence, we introduce CGOcc, a supervised multi-camera framework that explicitly models Cross-Gaussian dependencies for 3D semantic Occupancy prediction. CGOcc integrates appearance and geometric observations to refine Gaussian states and encodes cross-Gaussian dependencies in a diagonal-plus-low-rank representation. To incorporate these dependencies into voxel classification, CGOcc uses a class-conditional projection to extract diagonal and signed cross-Gaussian statistics without constructing a dense dependency matrix. The readout then refines occupied-class probabilities using these statistics while preserving a learned reference's occupied mass and free probability. Extensive experiments on nuScenes demonstrate CGOcc achieves state-of-the-art performance, with larger gains at high Gaussian overlap and improved predictions near boundaries. Ablation studies validate the effectiveness of covariance-aware prediction and support its synergy with geometric interaction.

cs.CV↗

A Survey on Physics-based Differentiable Rendering

Physics-based differentiable rendering has emerged as a powerful technique in computer graphics and vision, with a broad range of applications in solving inverse rendering tasks. At its core, differentiable rendering enables the computation of gradients with respect to scene parameters, allowing optimization-based approaches to solve various problems. Over the past few years, significant advancements have been made in both the underlying theory and the practical implementations of differentiable rendering algorithms. In this report, we provide a comprehensive overview of the current state of the art in physics-based differentiable rendering, focusing on recent advances in general differentiable rendering theory, Monte Carlo sampling strategy, and computational efficiency.

cs.GR↗

Fast Learning Radiance Fields by Shooting Much Fewer Rays

Learning radiance fields has shown remarkable results for novel view synthesis. The learning procedure usually costs lots of time, which motivates the latest methods to speed up the learning procedure by learning without neural networks or using more efficient data structures. However, these specially designed approaches do not work for most of radiance fields based methods. To resolve this issue, we introduce a general strategy to speed up the learning procedure for almost all radiance fields based methods. Our key idea is to reduce the redundancy by shooting much fewer rays in the multi-view volume rendering procedure which is the base for almost all radiance fields based methods. We find that shooting rays at pixels with dramatic color change not only significantly reduces the training burden but also barely affects the accuracy of the learned radiance fields. In addition, we also adaptively subdivide each view into a quadtree according to the average rendering error in each node in the tree, which makes us dynamically shoot more rays in more complex regions with larger rendering error. We evaluate our method with different radiance fields based methods under the widely used benchmarks. Experimental results show that our method achieves comparable accuracy to the state-of-the-art with much faster training.

cs.CV↗