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

ArcFace由2019年邓建康等人提出，采用加性角度间隔损失

1

Sources

50%

Confidence

Long-term

Relevance

8/7/2026

First Seen

Sources

AI人脸识别不准怎么办？六大实战优化方案详解

Related Claims

UnverifiedArcFace proposed the Additive Angular Margin Loss in 2018, achieving over 99.8% recognition accuracy on standard benchmarks like LFW77% similar UnverifiedArcFace是专为人脸识别优化的深度度量学习方法，通过引入加性角度间隔损失函数，在人脸身份特征提取精度上优于通用视觉编码器72% similar Unverified作者设计损失函数直接计算生成图像的人脸嵌入与目标人脸嵌入之间的距离，实现端到端可微的人脸相似度优化56% similar Unverifiedoutline支持outline-offset属性，可以设置负值让轮廓向内收缩53% similar UnverifiedHorizonSteady模式可在画面倾斜较大角度时保持水平，但会裁切画面、损失一定视角，介意广角损失的需权衡53% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/705629

API

curl https://kongchang.com/api/v1/knowledge/claims/705629

MCP

get_claim(id=705629)