Time-varying implicit neural representations (INRs) provide a compact representation of scientific volumes and, for modalities such as dynamic X-ray computed tomography (CT), are often the only practical way to represent the data. However, interactive volume rendering of INRs is challenging, as cheap memory lookups are replaced by expensive neural inferences, hindering the performance.
The paper introduces SVRecon, a generalizable neural surface reconstruction framework that uses sparse volumetric representations to achieve high-resolution 3D reconstruction. It employs a two-stage architecture: first predicting occupied voxels with an occupancy network, then rendering only within those regions using specialized sparse algorithms. This approach allows reconstruction at resolutions up to 512³ on 32 GB hardware, producing smoother and more precise surfaces, especially in sparse-view scenarios.
By Aoxiang Fan, Corentin Dumery, Nicolas Talabot, Ming Xu, Hieu Le, Pascal Fua
DiffusionShadow introduces a diffusion-based shadow caching framework for neural volume rendering, compressing many pre‑computed shadow INRs into a single diffusion model conditioned on lighting direction. The method encodes shadow coefficient volumes as shadow INRs, trains the diffusion model to predict shadow INR weights at inference, and integrates directly with standard INR renderers without extra runtime sampling. Experiments demonstrate faster rendering than traditional approaches while avoiding the large storage overhead of independent INRs, producing shadows that closely match reference results.
By Kai-Chen Tung, Qi Wu, David Bauer, Mengjiao Han, Silvio Rizzi, Kwan-Liu Ma
arXiv:2607. 20970v1 Announce Type: new Abstract: Implicit neural representations (INRs) for time-varying volumetric data are typically trained using dense sampling over spatiotemporal coordinates, where each observation corresponds to a single point in space and time.
By Weihan Zhang, Xuan Zhao, Yenwen Peng, Yuqi Chen, Jun Tao
ReSplat introduces a recurrent Gaussian splatting model that iteratively refines 3D Gaussians using the rendering error as a feedback signal, avoiding explicit gradient computation. The method starts from a compact reconstruction in a subsampled space, producing far fewer Gaussians than prior per‑pixel models, which reduces computational cost. Experiments on multiple datasets, view counts, and resolutions show state‑of‑the‑art performance with faster rendering speeds.
By Haofei Xu, Daniel Barath, Andreas Geiger, Marc Pollefeys
arXiv:2607. 01164v1 Announce Type: new Abstract: Recent work has shown that implicit neural representations (INRs) can be trained to effectively compress structured and unstructured volume data, allowing for direct data querying with a reduced memory footprint.
By Landon Dyken, Sharmistha Chakrabarti, Nathan Debardeleben, Steve Petruzza, Qi Wu, Will Usher, Sidharth Kumar