Comments (7)
http://mmlab.ie.cuhk.edu.hk/archive/CNN_FacePoint.htm
用这个方法跑出来的landmark
from face_verification_experiment.
检测人脸上的五点的时候出现了,Warning:roi exceed the original image, 这是不是代表五点的坐标超出了原始图像尺寸?请教一下博主您当时有没有出现这种情况,如果有的话,您是怎么处理这些数据的。谢谢!
from face_verification_experiment.
直接跑就好了,我是跑完了以后人工检验的
from face_verification_experiment.
谢谢博主,alignment已经做好了,我现在用这些数据在训练vgg-16的网络,已经训练了3天了,但是分类错误率(错误率一直在百分之九十多)和损失函数一直不见下降,learning rate和batchsize我都调整过,也做了random crop依然没什么效果,博主您对这种情况有没有什么建议啊,另外您的这个模型跑了多久最终训练好得?特别感谢!
from face_verification_experiment.
额,注意仔细看vgg-16那篇论文,要先训练浅层网络,然后用浅层网络的结果初始化深层网络才比较容易收敛。我没试过那么深的网络,我觉得直接train可能比较难收敛,另外关于深层网络收敛难的问题可以关注MSRA最近关于Deep Residual Learning的一系列工作
from face_verification_experiment.
好的,非常感谢。
2016-03-21
blackbest
发件人:Alfred Xiang Wu [email protected]
发送时间:2016-03-21 10:45
主题:Re: [face_verification_experiment] 人脸上五个点的坐标是怎么得到的? (#35)
收件人:"AlfredXiangWu/face_verification_experiment"[email protected]
抄送:"Skywalkerfh"[email protected]
额,注意仔细看vgg-16那篇论文,要先训练浅层网络,然后用浅层网络的结果初始化深层网络才比较容易收敛。我没试过那么深的网络,我觉得直接train可能比较难收敛,另外关于深层网络收敛难的问题可以关注MSRA最近关于Deep Residual Learning的一系列工作
—
You are receiving this because you commented.
Reply to this email directly or view it on GitHub
from face_verification_experiment.
那么工具标不出或者标错的图片,就不用于训练吗?人工量好大呀:)
from face_verification_experiment.
Related Issues (20)
- There may be some errors in the 'labels' HOT 3
- Where do you get the 5 facial points for MSCELEB dataset. Do you use some other model to find the facial points ? HOT 1
- Xavier vs Gaussian HOT 1
- I feel the size of the trainning data doubt HOT 1
- no an issue but a doubt HOT 2
- about lfw labels in model C-LightenedCNN_C_lfw.mat HOT 3
- Error occurred when computing accuracy
- Query
- no such file in your clean list
- Eltwise层名字拼写错误有影响吗?
- MegaFace Evaluation: it generated 'OpenCV Error' when running experiment.py with own features HOT 1
- hello,your model A and B is small,why your model C is bigger than B by using the same cnn architecture HOT 1
- can't get the same result HOT 3
- How to use fine tuned model to do classification for an image in light CNN?
- Any architecture patent information HOT 1
- the issue about architecture
- LightenedCNN_A_deploy.prototxt
- The evaluation of CASIA-NIR2.0
- Request Pretrained lightenedCNN model in .mat format HOT 1
- Computation of rank 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from face_verification_experiment.