Unverified80% confidenceFactTime unknown
TensorRT内置了数百种融合模式的识别规则，能够自动发现并应用层融合优化
1
Sources
80%
Confidence
Long-term
Relevance
6/1/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe Viola-Jones face detection algorithm progressively filters candidate regions through a series of cascade classifiers ranging from simple to complex, and is a classic example of cascade strategy.62% similarUnverifiedTensorRT的内核自动调优机制会针对目标GPU的具体架构（如Ampere、Hopper等）对候选内核进行性能基准测试61% similarUnverifiedscikit-learn provides a unified fit/predict interface that encapsulates dozens of classical algorithms including linear regression, random forests, and SVM.60% similarUnverifiedSemgrep 采用基于模式匹配的轻量级方法，开发者可以用类似代码的语法编写检测规则60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/21234API
curl https://kongchang.com/api/v1/knowledge/claims/21234MCP
get_claim(id=21234)