Explore a comprehensive introduction to developing a complete pipeline for supervised image classification using convolutional neural networks (CNNs) in this 51-minute conference talk. Focus on classifying hand gestures as rock, paper, or scissors while learning about dataset acquisition, data splitting, CNN architectures, and model training and evaluation. Encounter common setbacks to understand key machine learning concepts such as overfitting, generalization ability, and dataset augmentation strategies. Discover spectacular applications of these techniques in robotics, industry, and biomedicine. Delivered through slides and a Python Jupyter notebook using Keras and TensorFlow, this conceptual and language-independent talk requires no prerequisites, making it accessible to beginners and experienced developers alike.