This chapter is a fundamental introduction to the central role of gait recognition in biometrics, particularly in the context of Artificial Intelligence (AI). It elucidates the importance of gait as a behavioral biometric modality, offering advantages such as contactless and unobtrusive data collection. Focusing on vision-based gait recognition using image sensors, the chapter delves further into the multifaceted applications of gait recognition, covering person identification, re-identification, and authentication. Challenges, referred to as covariant factors, such as viewing angles, walking speed, and environmental variables, are discussed, highlighting the adaptability required for robust gait recognition systems. Finally, the chapter introduces the concept of gait feature representation, contrasting model-based and model-free approaches, laying the groundwork for subsequent discussions on DL methods for gait recognition. The typical Convolutional Neural Network (CNN) architecture is presented as a stepping stone for a nuanced exploration of other architectures. The concepts covered lay the groundwork for advanced investigations in subsequent chapters, contributing to the construction of necessary knowledge in the field of gait recognition.