Kathmandu Valley 高精度土地利用分割数据集发布，合成数据增强反而降低精度
Fine-resolution land-use segmentation and synthetic image-label augmentation in the Kathmandu Valley
做遥感分割的朋友可以看看，Kathmandu Valley 这份 202 平方公里的六类标注数据集很难得，还顺手拆穿了随机切分的空间泄漏问题。更有意思的是扩散模型生成的数据反而拉低精度，做数据增强前真的要先过滤。
研究团队发布了一份手工标注的六类土地利用语料库，覆盖 Kathmandu Valley 202 平方公里、408 幅高分辨率 RGB 影像，填补了全球南方快速城市化城市地块级数据的空白。数据集难点突出：RGB 光谱相似度高、类别不平衡接近 100:1，河流等关键结构在标准层级解码器的预测步长下仅两三个像素宽。
技术层面，作者对五种分割架构进行评测，发现常规随机数据切分掩盖了大量空间泄漏——改用严格的序列块切分后 mIoU 最多下降 0.095，这一惩罚超过架构之间的差距。他们在 SegFormer 基线上加入半分辨率残差细节路径、边界可靠性重加权、拓扑感知损失和 presence 平衡采样器，仅增加 1.9% 参数就在所有测试折上提升 mIoU，增益集中在稀有细结构类别；但两个模型在外部 LoveDA 基准上成绩相同。
一个反直觉结论值得注意：用扩散模型生成的图像-标签对做增强，在同等算力下反而降低所有架构的精度，尤其伤害本想改善的稀有类别，说明盲目混入合成数据可能严重损害性能，训练前必须严格评估和过滤生成样本。
Fine-resolution land-use segmentation and synthetic image-label augmentation in the Kathmandu Valley
Parcel-scale land-use data is missing for most rapidly urbanizing Global South cities, as current segmentation architectures rely on multispectral imagery and planned layouts. We introduce a hand-annotated six-class land-use corpus covering 202 km$^2$ of Kathmandu Valley across 408 fine-resolution RGB scenes. The dataset is uniquely challenging: RGB spectral similarity is high, class imbalance approaches 100:1, and key structural features like rivers measure just two to three pixels wide at a standard hierarchical decoder's prediction stride. Evaluating five architectures reveals that standard random data splitting conceals massive spatial leakage. Switching to a strict sequence-block split drops performance by up to 0.095 mIoU. This penalty exceeds the performance gap between the architectures themselves. We modified the SegFormer baseline by integrating a half-resolution residual detail path, boundary-reliability reweighting, a topology-aware loss, and a presence-balanced sampler. Adding only 1.9% more parameters, the enhanced architecture improves mean IoU across every test fold, with gains concentrated in rare, thin-structure classes. However, both models achieve identical results on the external LoveDA benchmark. Finally, augmenting with diffusion-generated pairs at matched compute degraded accuracy across all architectures, especially harming rare classes it aimed to improve. This demonstrates that blindly incorporating synthetic data can severely damage performance, highlighting a critical need to rigorously evaluate and filter generated samples before training.
This report is categorized under AITOP's "Papers" desk. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include SegFormer, 遥感分割, LoveDA, 语义分割.
Source attribution: curated from arXiv cs.AI (@Pranil Parajuli, Sugam Dahal, Kiran Chandra Dahal).