Papers in 100 Lines of Code·Sep 21AnySplat Turns 32 Unposed Photos Into a 3D Scene in 1.4 SecondsBuilt on VGGT and trained without 3D ground truth, it predicts cameras, depth, and millions of 3D Gaussians in one forward pass — and…A response icon1A response icon1
Papers in 100 Lines of Code·Sep 11Apple Turns One Photo Into 1.2 Million 3D GaussiansSHARP reconstructs a metric-scale 3D scene in under a second, using a single forward pass instead of diffusion or per-scene optimization.A response icon1A response icon1
Papers in 100 Lines of Code·Aug 273D Gaussian Splatting No Longer Has to Start From ScratchFeed-Forward 3D Gaussian Splatting: From Proof of Concept to ProductionA response icon1A response icon1
InHTX DSAIbyAdrian Law·Jul 28From Photos to Playable Mesh: A Technical Primer on Modern 3D Scene ReconstructionHow structure-from-motion and mesh-based Gaussian splatting turn an ordinary folder of photos into a real-time, exportable 3D scene
Fivemon·Jun 23Tiled-MRPNN — real-time photorealistic light transport in participating mediaHistorically, real-time light transport in participating media has mainly been handled with simplified physics simulations and crude…
Shanmuka Sadhu·Jun 9Paper Insights: DIFF4SPLAT: Repurposing Video Diffusion Models for Dynamic Scene GenerationI recently came across a GitHub repository with many papers on physically plausible video generation, and I found this paper. This paper…
Papers in 100 Lines of Code·Dec 29, 2025What Happens If You Initialize 3D Gaussian Splatting From a Very Sparse Point Cloud?As you might know, most 3D Gaussian Splatting (3DGS) pipelines are initialized from a point cloud-typically produced by COLMAP/SfM. When…
Papers in 100 Lines of Code·Dec 22, 2025Extracting Depth From Any Pretrained 3DGS (No Retraining Needed)Extract depth maps from any pretrained 3D Gaussian Splatting without retraining, depth loss, or special…
Papers in 100 Lines of Code·Dec 19, 2025Accelerating 3D Gaussian Splatting With Speedy Splat3D Gaussian Splatting (3DGS) is one of the fastest rendering methods in neural graphics, but the original algorithm assigns far too many…
