欢迎光临散文网 会员登陆 & 注册

斯坦福李飞飞终于把【计算机视觉】给讲明白了!全程16集,学不会我退出IT界!深度

2022-12-15 10:13 作者:呆毛菜  | 我要投稿


1.第1讲_卷积简介. P1 - 38:33


ImageNet Challenge Winning algorithms

2012: AlexNet: 7 layer

2014: GoogleNet/VGG: 19 layers

2015: Microsoft Research Asia: ResNet 152 layers

deeper and deeper with tuning


1.第1讲_卷积简介. P1 - 40:53


Invention from 1998 in Bell Lab: LeCun et al.

  • for handwritten checks

Grown Popularity in 2012:

  • GPU computation
  • data-hungry: deploy of PASCAL & ImageNet


1.第1讲_卷积简介. P1 - 45:06


3D understanding of the world

  • activity
  • AR/VR


1.第1讲_卷积简介. P1 - 46:21


EX. Visual Genome

  • more connections

EX. Fei-Fei's Gradual work

  • description of photo
  • from external knowledge


1.第1讲_卷积简介. P1 - 50:23


People: 3 Instructor, 18 TA, graduated PhD

Textbook: Deep Learning

  • by Goodfellow, Yoshua Bengio, Aaron Courville
  • free online


1.第1讲_卷积简介. P1 - 53:13


thorough & detailed: from mechanics to train

pratical

state-of-the-art

fun

斯坦福李飞飞终于把【计算机视觉】给讲明白了!全程16集,学不会我退出IT界!深度的评论 (共 条)

分享到微博请遵守国家法律