Topodpo improves aerial image segmentation accuracy and structure alignment
TopoDPO: Dense Preference Alignment for Structure-Sensitive Aerial Referring Segmentation
Computer Vision and Pattern RecognitionArtificial Intelligence
Summary
Segmenting objects accurately in aerial images is tricky because different parts of the object need different kinds of attention. The authors present TopoDPO, a new method that better matches these local differences by carefully aligning preferences in the segmentation. Their approach helps capture the shape and structure of objects more precisely, especially when traditional methods struggle with complex shapes. This leads to small but measurable improvements in segmentation quality on a large test set of aerial images.
What this means in practice
- •For geospatial analysts: Segment aerial imagery with improved accuracy and preserved structural details for mapping and monitoring land use.
- •For remote sensing software developers: Integrate TopoDPO to enhance segmentation models in processing satellite or drone images showing complex object boundaries.
Authors
Huaiyuan Peng, Jisheng Dang, Wencan Zhang, Jingwen Zhao, Hong Peng, Wenjun Ma, Bin Hu, Qi Tian, Tat-Seng Chua
Abstract
Preference optimization for multimodal segmentation faces a granularity mismatch: dense masks contain spatially heterogeneous preference signals, whereas image-level objectives assign one shared nonlinear weight to all spatial margins. We introduce TopoDPO, a topology-aware dense preference alignment framework for structure-sensitive aerial referring segmentation. Topo-Prior mines informative preference pairs using connected-component discrepancy and macro-corner recall, while the dense objective applies the preference penalty to each local margin before spatial aggregation against a frozen supervised reference. On Topo-RS-Hard, a 4590-instance structural stress set, TopoDPO improves intersection over union (IoU) from 50.91% to 51.24% over image-level direct preference optimization (DPO) and reduces Betti-0 error from 1.34 to 0.90; paired image-cluster bootstrap intervals against image-level DPO exclude zero for all five reported metrics. Matched supervised controls and a complete factorial analysis separate the effects of topology-aware pair mining, localized correction, and dense reference-relative preference weighting. An unfiltered stratification over all 39,244 validation records further shows the clearest structural gains on samples with moderate or severe pre-alignment component-count errors. The code is publicly available at: https://github.com/phy709/TopoDPO.