Unverified50% confidenceFactExact time
2012年前后基于深度学习的OCR已能以超过99%的准确率破解reCAPTCHA v1
1
Sources
50%
Confidence
Long-term
Relevance
7/12/2026
First Seen
Sources
反AI字体：人能读懂、机器识别不了的对抗性设计
hackernewshackernews7/11/2026
Related Claims
Unverified深度学习时代的LSTM和CNN架构大幅提升了OCR识别准确率，但真实业务场景中端到端准确率仍难以达到100%76% similarUnverifiedOpen-source OCR engines such as Tesseract and PaddleOCR have incorporated deep learning backends.63% similarUnverifiedReLU, defined as max(0, x), rapidly gained widespread adoption in the 2010s due to its computational efficiency and gradient stability, becoming the default activation function in modern deep learning.63% similarUnverified开源社区对o1的复现尝试包括DeepSeek-R1和QwQ，通过强化学习训练模型的推理过程是实现类似能力的关键路径62% similarUnverifiedDeepSeek employs FP8 mixed-precision quantization and Multi-head Latent Attention (MLA) and other technologies in its inference deployment.62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/491920API
curl https://kongchang.com/api/v1/knowledge/claims/491920MCP
get_claim(id=491920)