深度神经网络PPT培训课件
合集下载
相关主题
- 1、下载文档前请自行甄别文档内容的完整性,平台不提供额外的编辑、内容补充、找答案等附加服务。
- 2、"仅部分预览"的文档,不可在线预览部分如存在完整性等问题,可反馈申请退款(可完整预览的文档不适用该条件!)。
- 3、如文档侵犯您的权益,请联系客服反馈,我们会尽快为您处理(人工客服工作时间:9:00-18:30)。
Neural network Back propagation
Nature
1986
Deep belief net Science
Speech
2006
2011 deep learning results
• Solve general learning problems • Tied with biological system
Deep belief net Science
Speech
1986
2006
2011 2012
ImageNet 2013 – image classification challenge
MSRA, IBM, Adobe, NEC, Clarifai, Berkley, U. Tokyo, UCLA, UIUC, Toronto …. Top 20 groups all used deep learning
– Baidu
Neural network Back propagation
Deep belief net Science
Speech
Face recognition
1986
2006
2011 2012
2014
Deep learning achieves 99.53% face verification accuracy on Labeled Faces in the Wild (LFW), higher than human performance
深度神经网络
内容
深度神经网络发展历史、背景 动机——Why Deep Learning? 深度学习常用模型
Neural network Back propagation
Nature
1986
历史
• 解决了一般性学习问题 • 与生物系统相联系
Neural network Back propagation
nonlinearity, dropout) • New development of computer architectures
– GPU – Multi-core computer systems
• Large scale databases
Big Data !
深度学习浪潮
IT Companies are Racing into Deep Learning
– Goowenku.baidu.comle
• “on our test set we saw double the average precision when compared to other approaches we had tried. We acquired the rights to the technology and went full speed ahead adapting it to run at large scale on Google’s computers. We took cutting edge research straight out of an academic research lab and launched it, in just a little over six months.”
• ImageNet 2014 – object detection challenge
Neural network Back propagation
Deep belief net Science
Speech
1986
2006
2011 2012
• Google and Baidu announced their deep learning based visual search engines (2013)
But it is given up…
Neural network Back propagation
Nature
1986
Deep belief net Science
Speech
2006
2011 2012
Object recognition over 1,000,000 images and 1,000 categories (2 GPU)
Neural network Back propagation
Nature
1986
…… …… …… ……
历史
Deep belief net Neural networks
Science
is coming back!
2006
• Unsupervised & Layer-wised pre-training • Better designs for modeling and training (normalization,
• ImageNet 2013 – object detection challenge
Neural network Back propagation
Deep belief net Science
Speech
1986
2006
2011 2012
ImageNet 2014 – Image classification challenge
A. Krizhevsky, L. Sutskever, and G. E. Hinton, “ImageNet Classification with Deep Convolutional Neural Networks,” NIPS, 2012.
Neural network Back propagation
Y. Sun, X. Wang, and X. Tang. Deep Learning Face Representation by Joint IdentificationVerification. NIPS, 2014.
Y. Sun, X. Wang, and X. Tang. Deeply learned face representations are sparse, selective, and robust. CVPR, 2015.
Nature
1986
历史
w1 w2
w3
x1
x2
x3
Neural network Back propagation
Nature
1986
历史
2006
• 解决了一般性学习问题 • 与生物系统相联系
But it is given up…
• SVM • Boosting • Decision tree •…