Giter VIP home page Giter VIP logo

Comments (8)

Zehaos avatar Zehaos commented on May 28, 2024

说实话我不知道你这样做的意义哦,为什么不直接用mobilenet呢?

  1. hyper-params不好说,肯定要调的
  2. 你给的信息比较有限, loss震荡可能原因很多,可能是lr大了。

from mobilenet.

guanxiongsun avatar guanxiongsun commented on May 28, 2024

@Zehaos 因为我这个网络不是用于分类,用作人体关键点检测,这个网络其实是一个级联的结构,分为多个stage,每个stage中包含7个convlayer,stage其实是一样的,只不过下一个stage会用到本次stage的结果,类似一个残差学习的过程。我这样做的目的是想看看mobileconv这种方式能否很好的逼近原始conv的结果,而且我觉得这样从后向前改动,比直接把所有的conv全部改成mobileconv的形式容易训练。这个网络的底座是vgg,之后级联3个stage。您觉得我是可以直接把vgg改成mobilenet,然后把所有stage都改成mobileconv直接scratch训练吗?

from mobilenet.

Zehaos avatar Zehaos commented on May 28, 2024

@sunguanxiong 噢噢,mobilenet网络设计上跟vgg是一个套路,应该可以直接把vgg换成mobilenet。
另外,关键点检测需要比较fine的特征,你可以加一些skip-connection或者deconv(或者直接跑多几个stage)。

from mobilenet.

Zehaos avatar Zehaos commented on May 28, 2024

@sunguanxiong 有tf版本的realtime pose estimation?

from mobilenet.

guanxiongsun avatar guanxiongsun commented on May 28, 2024

@Zehaos 我用的caffe版本的,mobilenet也是caffe版本的

from mobilenet.

Sunmyfong avatar Sunmyfong commented on May 28, 2024

刚好在 YOLO 上做了几个实验:
1.将 tiny-yolo-voc 前10层保持不变,后面的卷积用 depthwise conv 替换,训练 6W 轮,batchsize 64;
Recall 下降 6%,卷积层速度提升:CPU 8x 左右,GPU 感觉不到提升。模型9M,原64M。
2.将tiny-yolo-voc 卷积层全部换成DWConv,没pre-trained model,直接训练,LOSS一直震荡,
看到生成的model 才 800K,遂弃。

from mobilenet.

austingg avatar austingg commented on May 28, 2024

@Sunmyfong detection 这些应用,还是要用pretrained model才行

from mobilenet.

Zehaos avatar Zehaos commented on May 28, 2024

@Sunmyfong 可以先小lr warm up 一两个epoch,再train,不会震荡就是好事情。

from mobilenet.

Related Issues (20)

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo 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.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.