Populate Byline Age with airSlate SignNow

Populate byline age

[Music] hi guys welcome back to data everyday 200th video today we are looking at a data set of images of people and we're going to try to predict the age from an image so normally when you're using image data you're doing some sort of image classification or object detection but today we're actually going to be doing a regression task based on the images so we're going to use the name of the folder as an uh integer value on a continuous scale for predicting the age of the person in a given image so let's hop into the notebook we're going to be using numpy pandas the path object from path lib and os.path for working with the data for pre-processing we'll use the train test split function from sklearn then for building our image data generators and the model itself we'll be using tensorflow and we'll evaluate the performance of the model with the r squared score so let's go ahead and import all of that and the first thing i want to do is get the name of the folder that we're using that contains all of our images so we're going to be using this 20 to 50 data set just because it has fewer images that we have to load up so we're going to be trying to predict the age of people ages 20 to 50. now i'm not going to worry about the train test split they gave us already i'm going to use all the images both in train and test so i'm just going to use this as the folder 20 to 50 here so i'm going to call this image der and uh there we go now i don't want to leave it just as a string i'm actually going to turn it into a path object from pathlib which allows us to do some nice operations on it then we'll go ahead and create our file data frame so that the file data frame is what we're going to use to flow the images from the disk into our model we're going to use flow from data frame from the keras image data generator and the file data frame just has to contain one column with the file paths and one column with the target values associated with each image so if we take imager which is a path lib path object and use the glob function we can actually use glob expressions to find images in the directory so i'm going to use the expression star star star.jpg once this is saying is anything you find regardless of directories this can contain subdirectories just as long as the final result is something.jpg so this is essentially saying find all images recursively through the whole directory sub tree so if we run this it's actually just a generator object so i'm going to turn it into a list to get back the list of file paths and...