Research map / Point clouds / Shape and surface completion
Underground Plant Exploration: Non-Destructive 3D Root Assessment with GPR Based on Point Graph Neural Network
Yuwei Zhou, Guoyu Lu
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 MobileNetV2-based hyperbola detection network instead of SE-SSD/Faster R-CNN/DiffusionDet | root hyperbola detection · Simulated 2D GPR root detection | AP: 0.857 vs 0.840 (best baseline) | Improved |
| use point graph network with attention and upsampling for 3D root reconstruction | root shape reconstruction · 3D root reconstruction (5 test roots, mean) | EMD (x100): 5.03 vs 7.08 / 7.57 / 7.12 | Improved |
| use point graph network with attention and upsampling for 3D root reconstruction | root shape reconstruction · 3D root reconstruction (5 test roots, mean) | CD (x100): 2.03 vs 2.90 / 3.29 / 3.00 | Improved |
| add the point graph network (ablation) | root shape reconstruction network · Synthetic GPR root reconstruction | CD and EMD | Improved |
| add the upsampling module (ablation) | root shape reconstruction network · Synthetic GPR root reconstruction | CD and EMD | Improved |
| add point feature attention (ablation) | root shape reconstruction network · Synthetic GPR root reconstruction | CD and EMD | Improved |
| use compact detection network (MobileNetV2) integrated into reconstruction pipeline | full pipeline · Model size | parameters: 20.98M vs 63.48M (Feng et al.) | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- SE-SSD — SE-SSD: Self-Ensembling Single-Stage Object Detector From Point Cloud (CVPR 2021) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- EDENet: Echo Direction Encoding Network for Place Recognition Based on Ground Penetrating Radar (AAAI 2025 · similarity 0.82)
- Dual-branch Graph Feature Learning for NLOS Imaging (AAAI 2025 · similarity 0.78)
- GridFormer: Point-Grid Transformer for Surface Reconstruction (AAAI 2024 · similarity 0.77)
- Towards Automated Petrography (NeurIPS 2025 · similarity 0.77)
- Auto-Linear Phenomenon in Subsurface Imaging (ICML 2024 · similarity 0.77)
- SGNet: Structure Guided Network via Gradient-Frequency Awareness for Depth Map Super-resolution (AAAI 2024 · similarity 0.77)
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.
