Feed Entities Podcasts Deep Dives Developers About

[KongchangAI]

Real-time AI knowledge base. Every claim is traceable to its source, with data freshness under 24 hours.

Powered by 240+ first-hand AI sources

Knowledge

Feed
Hot Entities
Podcasts

Developers

CLI
MCP Server
REST API
About

© 2026 KongchangAI. The freshest, most accurate AI knowledge

Unverified50% confidenceFactExact time

测量协方差矩阵 R 越大滤波器越不信任该传感器读数，越小则越快响应观测变化

1

Sources

50%

Confidence

Long-term

Relevance

7/12/2026

First Seen

Sources

多传感器卡尔曼融合实战：无人机追踪系统完整解析

redditr/learnmachinelearning7/11/2026

Related Claims

Unverified为每个传感器独立建模测量协方差矩阵R是多传感器融合区别于单传感器扩展的本质，R设置过小会导致过度信任该传感器测量值78% similar Unverified测量精度越高的探头响应速度往往越慢，追求快速反应的孵化场景与追求高精度稳定读数存在权衡，孵化建议优先响应速度69% similar UnverifiedTwo-stage检测器先通过RPN生成候选框再逐一分类，精度高但速度慢；One-stage检测器将候选框生成与分类合并为一步，牺牲少量精度换取数十倍速度提升68% similar Unverified当标注不一致存在时，增加更多数据只会成倍放大混乱，向模型提供自相矛盾的监督信号67% similar Unverified量化误差具有非均匀分布特性，注意力机制的Q/K/V矩阵和最后几层对量化更为敏感66% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/490365

API

curl https://kongchang.com/api/v1/knowledge/claims/490365

MCP

get_claim(id=490365)