SG-AMP: Scene-Graph-Guided Active Perception and Semantics-Aware Motion Planning for Pepper Plants
2026-09-01 • Robotics
Robotics
AI summaryⓘ
The authors introduce SG-AMP, a system that helps robots better understand and explore plant scenes by combining depth sensing, mapping, and reasoning about plant parts like peppers and their stems. Their approach guesses where unseen parts might be and plans camera movements to check those spots for new information. Experiments show their method improves the accuracy of recognizing and mapping plant parts and also provides better measures of uncertainty in depth sensing. This could help robots navigate and interact with complex plant structures more effectively.
depth completionuncertainty estimationpanoptic mappingscene graphactive view planningsemantic segmentationinformation gainpepper pedunclerobotic perception
Authors
Rohit Menon, Shiva Rudra Lolla, Niklas Mueller-Goldingen, Gokul Chenchani, Ribana Roscher, Maren Bennewitz
Abstract
We present SG-AMP, integrating robust depth completion with input-conditioned uncertainty, persistent panoptic mapping, plant scene-graph reasoning, and semantics-aware active view-motion planning. Beyond inspecting uncertain observed regions, the scene graph explicitly hypothesizes unobserved pepper--peduncle attachments and directs close-range sensing toward them. Candidate views are selected according to expected information gain, while class-dependent motion costs distinguish protected peppers, peduncles, and stems from conditionally traversable foliage. On pepper data, the perception network achieves $55.27\%$ semantic mIoU, $38.67\%$ PQ, and $40.62\,\mathrm{mm}$ depth RMSE, while input-conditioned uncertainty improves NYUv2 NLL from $-1.6518$ to $-1.6925$ and AUSE from $0.0102$ to $0.0087$.