OpenCV 4 with Python Blueprints - Second Edition

Processing the dataset

Analogous to Chapter 7, Learning to Recognize Traffic Signs, we write a new dataset parser in data/emotions.py that will parse our self-assembled training set.

We define a load_data function that will load the training data and return a tuple of collected data and their corresponding labels, as follows:

def load_collected_data(path):    data, targets = [], []    with open(path, 'r', newline='') as infile:        reader = csv.reader(infile)        for label, sample in reader:            targets.append(label)            data.append(json.loads(sample))    return data, targets

This code, similar to all the processing codes, is self-contained and resides in the data/process.py file, similar to Chapter 7, Learning to Recognize Traffic Signs.

Our featurization function ...