English(EN) Continuous Ground-Truth Construction and a Recovery Policy for Air--Water Robotic Tracking

新策略改进水-空界面机器人追踪

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-09 04:00

研究人员开发了一种用于构建地面真实性数据和水-空界面机器人追踪恢复策略的新方法。地面真实性构建流程将运动捕捉数据与摄像机帧同步，校正水下投影，并包括手动审查，生成了包含22,346帧的测试集。跨介质恢复策略（CMRP）通过使用置信度触发的模板选择来增强追踪，提高了MixFormerV2的追踪覆盖率并降低了恢复延迟。 AI

影响提高了机器人追踪系统在挑战性环境中的鲁棒性。

排序理由这是一篇详细介绍针对特定技术问题的新方法和策略的研究论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 18 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 这是一篇详细介绍针对特定技术问题的新方法和策略的研究论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, other; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.AI TIER_1 English(EN) · Jiangong Xiao, Zhe Sun, Kanzhong Yao, Yuanbo Bi, Haofei Zhao, Ruixuan Hu, Guan Huang, Xuelong Li · 2026-10-09 04:00

面向水-气机器人追踪的连续真值构建与恢复策略

arXiv:2610.11096v1 Announce Type: cross Abstract: Visual tracking across the air-water interface is challenged by splashes, bubbles, refraction, reflections, and abrupt appearance changes that can temporarily invalidate observations. This setting poses two coupled difficulties: f…

本文如何被排名

报道来源 [1]

面向水-气机器人追踪的连续真值构建与恢复策略

相关实体

相关话题