Unverified50% confidenceEventExact time
The project built a proprietary dataset of approximately 3 million frames through dual-modal RGB and LWIR collaborative annotation combined with semi-automated data mining.
1
Sources
50%
Confidence
Long-term
Relevance
8/9/2026
First Seen
Sources
自建多模态数据集：从500帧到300万帧的自动化标注实践
Related Claims
Unverified作者通过双模态协同标注与半自动化数据挖掘流程，构建了规模约300万帧的自有数据集75% similarUnverifiedClickHouse等OLAP列式存储数据库通过将同列数据物理连续存储配合向量化执行引擎，可对万亿行数据实现亚秒级聚合查询62% similarUnverifiedThe project is based on the Lyra/Gyra framework and has accumulated hundreds of thousands of lines of C++ code and thousands of Blueprint assets used solely for data configuration.61% similarUnverified数据版本管理和特征存储等工程实践催生了DVC、MLflow、Feast等专业工具生态60% similarUnverifiedAnaconda is an integrated distribution packaged specifically for data science scenarios, pre-installed with NumPy, pandas, Jupyter, and others, with a size of approximately 3 GB.60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/714113API
curl https://kongchang.com/api/v1/knowledge/claims/714113MCP
get_claim(id=714113)