pil example pa1 dataloader (snippet)

Python pil example 'pa1 dataloader'

Functions in program:

def display_face(img):
def load_data(data_dir="./CAFE/"):

Modules used in program:

import numpy as np

python pa1 dataloader

Python pil example: pa1 dataloader

################################################################################
# CSE 253: Programming Assignment 1
# Code snippet by Jenny Hamer
# Winter 2019
################################################################################
# We've provided you with the dataset in CAFE.tar.gz. To uncompress, use:
# tar -xzvf CAFE.tar.gz
################################################################################
# To install PIL, refer to the instructions for your system:
# https://pillow.readthedocs.io/en/5.2.x/installation.html
################################################################################
# If you don't have NumPy installed, please use the instructions here:
# https://scipy.org/install.html
################################################################################

from os import listdir
from PIL import Image
import numpy as np


# The relative path to your CAFE-Gamma dataset
data_dir = "./CAFE/"

# Dictionary of semantic "label" to emotions
emotion_dict = {"h": "happy", "ht": "happy with teeth", "m": "maudlin",
    "s": "surprise", "f": "fear", "a": "anger", "d": "disgust", "n": "neutral"}


def load_data(data_dir="./CAFE/"):
    """ Load all PGM images stored in your data directory into a list of NumPy
    arrays with a list of corresponding labels.

    Args:
        data_dir: The relative filepath to the CAFE dataset.
    Returns:
        images: A list containing every image in CAFE as an array.
        labels: A list of the corresponding labels (filenames) for each image.
    """
    # Get the list of image file names
    all_files = listdir(data_dir)

    # Store the images as arrays and their labels in two lists
    images = []
    labels = []

    for file in all_files:
        # Load in the files as PIL images and convert to NumPy arrays
        img = Image.open(data_dir + file)
        images.append(np.array(img))
        labels.append(file)

    print("Total number of images:", len(images), "and labels:", len(labels))

    return images, labels



def display_face(img):
    """ Display the input image and optionally save as a PNG.

    Args:
        img: The NumPy array or image to display

    Returns: None
    """
    # Convert img to PIL Image object (if it's an ndarray)
    if type(img) == np.ndarray:
        print("Converting from array to PIL Image")
        img = Image.fromarray(img)

    # Display the image
    img.show()

Python links

Learn Python: https://pythonbasics.org/
Python Tutorial: https://pythonprogramminglanguage.com