Frame-to-Panorama 定位与上下文感知采样：用 220 张标注图训练船舶检测模型
Frame-to-Panorama Localization and Context-Aware Sampling for Scene-Specific Ship Detection in a Smart Marina Testbed
一篇很实用的数据筛选流水线：40718 帧 maritime 视频只挑 220 张标注，微调 YOLO26-m 就能到 AP50 94.78%，标注成本省了 99.5%。
这篇 arXiv 论文提出一套针对海事 PTZ 视频的数据筛选流水线，用 SuperPoint 和 LightGlue 将历史视频帧定位到参考全景图上，并加入天气与太阳状态元数据。
第二条上下文感知阶段通过 tile 级视觉嵌入和 Gaussian Mixture Model 聚类，补充地平线附近被低估的远距船舶样本。在 CMMI MDigi-I Smart Marina 测试床中，该流水线将 40,718 帧候选图像压缩到 220 张用于标注，降幅 99.
5%。在该子集上微调的 YOLO26-m 检测器在序列分组五折交叉验证下取得 AP50 94.
78%±0.51%、AP50-95 75.
10%±1.73%。
Frame-to-Panorama Localization and Context-Aware Sampling for Scene-Specific Ship Detection in a Smart Marina Testbed
Smart maritime infrastructures provide continuous access to heterogeneous sensing streams, enabling repeated experimentation, digital-twin development, and AI-based maritime services. However, sensing hardware alone is not sufficient for scene-specific model development: historical video streams must also be spatially indexed, contextualized, and reduced to informative subsets for annotation. This paper presents a frame-to-panorama localization and context-aware sampling pipeline for ship detection in historical PTZ maritime video lacking reliable pan, tilt, and zoom metadata. The main contribution is an end-to-end data-curation approach that recovers camera-view information from historical PTZ video and combines it with environmental context and visual diversity to construct compact, scene-specific training sets. Specifically, frames are localized on a reference panorama using SuperPoint and LightGlue, enriched with weather and solar-state metadata, and selected through diversity sampling to preserve variation across camera view and environmental conditions. A second context-aware stage targets under-represented distant-vessel cases near the horizon using tile-level visual embeddings and Gaussian Mixture Model clustering. Applied within the CMMI MDigi-I Smart Marina testbed, the proposed pipeline reduces 40,718 candidate frames to 220 images for annotation, corresponding to a 99.5% reduction. A YOLO26-m detector fine-tuned on this subset achieves a mean AP50 of 94.78% $\pm$ 0.51% and a mean AP50-95 of 75.10% $\pm$ 1.73% under sequence-grouped five-fold cross-validation. These results demonstrate that highly redundant infrastructure video streams can be transformed into compact, spatially and contextually diverse training sets for scene-specific detector adaptation while substantially reducing annotation effort.
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 SuperPoint, LightGlue, YOLO26, data labeling.
Source attribution: curated from arXiv cs.AI (@Ignat Romanov, Andreas Hadjipieris, Neofytos Dimitriou).
