Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

关于文本不同 导致的 训练精度差异的问题 #45

Closed
wangshuang-jiayou opened this issue Dec 18, 2023 · 1 comment
Closed

Comments

@wangshuang-jiayou
Copy link

wangshuang-jiayou commented Dec 18, 2023

您好!我在利用自己的数据作训练的过程中发现了一个问题:比如当文本是truck . truck mixer . heavy truck;再比如文本是insulator . dirty insulator . damadge insulator等,这种多类别包含了相同词汇的文本时,得到的预测结果有很多是 truck truck mixer、insulator dirty insulator等。然后我改变了类别的定义,比如说truck . concrete mixer . heavy让它们不再包含相同词汇,识别率会提升很多。

起初我以为是模型对某两个类别的特征区分能力比较差 导致它认为某物体会同时是这两个物体。后来我想了下,跟文本特征提取模块也有关系吧?像yolo这种没有文本特征提取分支的模型,相同的训练和验证集识别率就相对高一点

@BIGBALLON
Copy link
Collaborator

BIGBALLON commented Dec 23, 2023

I think this is the case because GDINO contains the BERT text branch, so text with no ambiguity or more distinction will bring better results.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants