dlcv_for_beginners

Why we included this project

These are the runnable examples that accompany a published Chinese textbook on deep learning and computer vision, organized chapter by chapter so beginners can watch each concept turn into working code. Working through it covers MNIST classification, image regression, transfer learning from pretrained ImageNet models, Siamese networks, and a small object-detection labeling tool. A bonus folder branches into adversarial samples, a simple 2D GAN, model fusion, and U-Net style segmentation in PyTorch. The author is honest that the later chapters and frameworks like Caffe have aged and that the repo is no longer updated, so it is best treated as a study companion and a historical snapshot of how these topics were taught, not as a base for production work.