Speech recognition, with speech as the research object, realizes natural language interaction between humans and machines. It is a branch of pattern recognition and involves many fields, such as physiology, psychology, linguistics, computer science, and signal processing. This chapter covers the evolution of speech recognition technologies, from early isolated word recognition systems to modern continuous speech recognition models. Classic algorithms such as Hidden Markov Models (HMM), Gaussian Mixture Models (GMM), and recent deep learning methods like DNN, LSTM, and BiLSTM are introduced, alongside cutting-edge end-to-end models such as CTC and Seq2Seq. This chapter also discusses challenges like robustness, cocktail party effects, and personalized recognition, and explores the latest developments such as hybrid network conformers and DFCNN models. Through practical applications and examples, this chapter illustrates the current state and future potential of speech recognition technology.