Research map / Architectures and efficiency / RNNs and normalization
Parallel and Scalable Heat Methods for Geodesic Distance Computation
Jiong Tao, Juyong Zhang, Bailin Deng, Zheng Fang, Yue Peng, Ying He
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use edge-based gradient formulation instead of face-based | parallel heat method · WelshDragon | peak memory (MB): 9,420 -> 4,819 | Improved |
| use edge-based formulation instead of face-based | parallel heat method · WelshDragon | mean relative error (%): 0.50% -> 1.00% | Hurt |
| use edge-based formulation instead of face-based | parallel heat method · MalteseFalcon | time (s): 834.52 -> 1283.56 | Mixed |
| replace Cholesky direct heat diffusion with parallel Gauss-Seidel | heat diffusion step · 63208-vertex model | time (s): 1.671s -> 0.169s | Improved |
| replace VTP exact geodesic solver with face-based ADMM heat method | geodesic distance computation · Lucy | time (s): 1320.15 -> 148.65 | Improved |
| replace direct-solver heat method with face-based ADMM method on very large meshes | geodesic distance computation · Cyvasse to ThunderCrab (39M-98M vertices) | feasibility: Out of Memory -> 355.42 | Improved |
| add global cycle integrability constraints for non-zero genus (ablation) | face-based method · genus-eight surface | mean relative error | No clear effect |
| add path constraints between multiple sources (ablation) | face-based method · multiple sources | mean relative error | No clear effect |
| use intrinsic Delaunay triangulation (ablation) | edge-based method · poorly triangulated meshes | mean relative error | Improved |
| increase thread count from one to eight | face-based method · parallelization test (600 GS iterations) | time | Improved |
| use Gauss-Seidel and ADMM instead of conjugate gradient solver | heat method · accuracy versus time | solution accuracy | Improved |
Closest papers by meaning
- Parallelised Differentiable Straightest Geodesics for 3D Meshes (CVPR 2026 · similarity 0.89)
- NeuroGF: A Neural Representation for Fast Geodesic Distance and Path Queries (NeurIPS 2023 · similarity 0.85)
- Efficient, Sparse Representation of Manifold Distance Matrices for Classical Scaling (CVPR 2018 · similarity 0.84)
- HotSpot: Signed Distance Function Optimization with an Asymptotically Sufficient Condition (CVPR 2025 · similarity 0.82)
- An Intrinsic Vector Heat Network (ICML 2024 · similarity 0.82)
- De-coupled NeuroGF for Shortest Path Distance Approximations on Large Terrain Graphs (ICML 2025 · similarity 0.82)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
