全新训练det模型要多少张图?
#14775
Replies: 1 comment
-
全新训练一个文本检测(det)模型需要的图片数量取决于多个因素,包括数据的复杂度、模型架构以及训练目标。 从提供的讨论来看,用户使用了300张图片进行训练,每张图片包含一个字符串,并进行了500轮训练。然而,训练结果显示模型在某些情况下仍然检测不出文本,说明数据量可能不足。 影响数据需求的因素:
推荐数据量:
结论:当前300张图片通常不足以训练一个优秀的文本检测模型,建议尝试增加数据量(几千张以上)、增强数据多样性,并且如果可能的话,采用预训练模型或者teacher-student训练策略来提升效果。 你可以参考相关讨论:ch_v4 det训练后,完全检测不出来。 Response generated by 🤖 feifei-bot | chatgpt-4o-latest |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
全新训练det模型要多少张图?
Beta Was this translation helpful? Give feedback.
All reactions