top | item 15994919

Computer Vision by Andrew Ng – Lessons Learned

109 points| jonbaer | 8 years ago |kdnuggets.com | reply

11 comments

order
[+] nerfhammer|8 years ago|reply
Where do you get pretrained networks from? I looked for awhile but wasn't able to find any

Like if I wanted one of these: https://github.com/hwalsuklee/tensorflow-generative-model-co...

[+] Choco31415|8 years ago|reply
I’ll see if I can’t make a directory of various pretrained networks and collections. Expect a new reply in ~2 days.
[+] dandermotj|8 years ago|reply
Googling LeNet, Reset, AlexNet or other well known architectures will provide you with links to the network weights.
[+] arnioxux|8 years ago|reply
Anyone know if paying for the specialization is worth it? (Or for any other online mooc certificate in general?)
[+] cube2222|8 years ago|reply
Well, it gets you some meat to fill your CV/LinkedIn with and the prices aren't steep really with 50$ a month or so.

I think you also don't have access to programming assignments otherwise, which are very useful IMO, though you can probably gain access in other ways.

[+] ky738|8 years ago|reply
Loved his Coursera "Machine Learning" course. A good mix of maths and intuition.
[+] ReverseCold|8 years ago|reply
FWIW I didn't like it that much. I learned much more from reading about it and then trying to code things.

I do understand that his course also gets you to code yourself, but they give you "helper functions" that obviously won't be there when actually making something, so I'd rather write those myself.