visualize_data.py

import matplotlib.pyplot as plt
import numpy as np
import os
import torch
import pandas as pd

from dataset import resize, ImageStandardizer, DogsDataset
from imageio import imread
from utils import config, denormalize_image

training_set = DogsDataset("train")
training_set.X = resize(training_set.X)
standardizer = ImageStandardizer()
standardizer.fit(training_set.X)

metadata = pd.read_csv(config("csv_file"))
metadata = metadata[metadata['partition'] != 'challenge'].reset_index(drop=True) 
print(
    "I will display some images. Click on the figure to refresh. Close the figure to exit."
)

N = 4
fig, axes = plt.subplots(nrows=2, ncols=N, figsize=(2 * N, 2 * 2))

pad = 3
axes[0, 0].annotate(
    "Original",
    xy=(0, 0.5),
    xytext=(-axes[0, 0].yaxis.labelpad - pad, 0),
    xycoords=axes[0, 0].yaxis.label,
    textcoords="offset points",
    size="large",
    ha="right",
    va="center",
    rotation="vertical",
)
axes[1, 0].annotate(
    "Preprocessed",
    xy=(0, 0.5),
    xytext=(-axes[1, 0].yaxis.labelpad - pad, 0),
    xycoords=axes[1, 0].yaxis.label,
    textcoords="offset points",
    size="large",
    ha="right",
    va="center",
    rotation="vertical",
)

for ax in axes.flatten():
    ax.set_xticks([])
    ax.set_yticks([])

while True:
    rand_idx = np.random.choice(np.arange(len(metadata)), size=N, replace=False)
    X, y = [], []
    for idx in rand_idx:
        filename = os.path.join(config("image_path"), metadata.loc[idx, "filename"])
        X.append(imread(filename))
        y.append(metadata.loc[idx, "semantic_label"])

    for i, (xi, yi) in enumerate(zip(X, y)):
        axes[0, i].imshow(xi)
        axes[0, i].set_title(yi)

    X_ = resize(np.array(X))
    X_ = standardizer.transform(X_)
    for i, (xi, yi) in enumerate(zip(X_, y)):
        axes[1, i].imshow(denormalize_image(xi), interpolation="bicubic")

    plt.draw()
    plt.savefig('img.png')
    if plt.waitforbuttonpress(0) == None:
        break

print("OK, bye!")