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

关于人脸框输出的问题 #16

Open
orliz opened this issue Jun 19, 2019 · 4 comments
Open

关于人脸框输出的问题 #16

orliz opened this issue Jun 19, 2019 · 4 comments

Comments

@orliz
Copy link

orliz commented Jun 19, 2019

含有人脸的图像输出,最后头像上有好几个框,不止一个框,只是为什么

@DefTruth
Copy link

DefTruth commented Aug 2, 2019

NMS算法价格处理重复框的参数,然后在detect_onet中设置filter_overlap=True,问题解决。
if filter_overlap: # 处理重复脸的问题 如果实际面积s<=inter,则说明被包含在大框中
inds = np.where(np.logical_and(ovr <= thresh, areas[order[1:]] > 2 * inter))[0]
else:
inds = np.where(ovr <= thresh)[0] # [1-d array] idx

@orliz
Copy link
Author

orliz commented Aug 19, 2019

NMS算法价格处理重复框的参数,然后在detect_onet中设置filter_overlap=True,问题解决。
if filter_overlap: # 处理重复脸的问题 如果实际面积s<=inter,则说明被包含在大框中
inds = np.where(np.logical_and(ovr <= thresh, areas[order[1:]] > 2 * inter))[0]
else:
inds = np.where(ovr <= thresh)[0] # [1-d array] idx

请问是加在py_nms()这个函数里嘛?

@DefTruth
Copy link

是的。不过这个要看你用的场景,如果只是关注大人脸,不考虑包含的情况可以这样改NMS。要是通用的目标检测就不一定适用,因为大的box包含小的box的情况还是比较常见的。另一种可能就是mtcnn的精度不够,够的话就不会把人脸的一小部分判断成人脸,换一个精度更高的模型吧,SSH/DSFD之类的,不用改原始NMS也基本不会有这个问题。我自己改完跑的MTCNN精度感觉都不够用。

@XUCharles
Copy link

NMS算法价格处理重复框的参数,然后在detect_onet中设置filter_overlap=True,问题解决。
if filter_overlap: # 处理重复脸的问题 如果实际面积s<=inter,则说明被包含在大框中
inds = np.where(np.logical_and(ovr <= thresh, areas[order[1:]] > 2 * inter))[0]
else:
inds = np.where(ovr <= thresh)[0] # [1-d array] idx

我还是没懂怎么修改, filter_overlap是什么

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

3 participants