These are chat archives for FreeCodeCamp/DataScience

27th
May 2018
Alice Jiang
@becausealice2
May 27 2018 00:59
Took me a second to figure out what this was. Thanks! @ndburrus :joy:
CamperBot
@camperbot
May 27 2018 00:59
becausealice2 sends brownie points to @ndburrus :sparkles: :thumbsup: :sparkles:
:star2: 2241 | @ndburrus |http://www.freecodecamp.org/ndburrus
Norvin Burrus
@ndburrus
May 27 2018 01:00
@becausealice2 you're welcome! enjoy :palm_tree:
Hèlen Grives
@mesmoiron
May 27 2018 05:10

Hey thanks @ndburrus something that I like too. I happen to be vaguely interested in cold cases. So slowly part of the pieces arecoming together.

that reminds me has anyone a good resource on boundary decection?say dividing images in blocks. I think I will need to be able to do that too. It’s actually a skill that humans use allthe time.

CamperBot
@camperbot
May 27 2018 05:10
mesmoiron sends brownie points to @ndburrus :sparkles: :thumbsup: :sparkles:
:star2: 2242 | @ndburrus |http://www.freecodecamp.org/ndburrus
Bigyan Karki
@bigyankarki
May 27 2018 16:30
@ndburrus thanks mate. will definitely look at it.
CamperBot
@camperbot
May 27 2018 16:30
bigyankarki sends brownie points to @ndburrus :sparkles: :thumbsup: :sparkles:
:star2: 2244 | @ndburrus |http://www.freecodecamp.org/ndburrus
Bigyan Karki
@bigyankarki
May 27 2018 17:29
@ndburrus so i looked into it. And, googles image dataset is simply awsome. But, i think its not fit for what I am trying to do.
I was trying to get photos of mine and three other friends from my Google photos account. And, make a dataset that will have four classes (for each person). And, train it through inception v3/v4 pretrained model by freezing all layers upto bottleneck layer, and change output layer to just four.
However, I am confused how my dataset should look it. Should it contain only faces? Or, whole body? and, how should it be structured? And, how should the dataset look like?
Bigyan Karki
@bigyankarki
May 27 2018 17:34
FYI I have already downloaded and sorted all the images from Google photos into four different folders.
Each class has about 100 instances. But, it can be increased through data augmentation. But, I don't know how the dataset should be structured.