// the find
Charmve/computer-vision-in-action
A computer vision closed-loop learning platform where code can be run interactively online. 学习闭环《计算机视觉实战演练：算法与应用》中文电子书、源码、读者交流社区（持续更新中 ...） 📘 在线电子书 https://charmve.github.io/computer-vision-in-action/ 👇项目主页
A single-author Chinese-language computer vision textbook packaged as a GitHub repo: notebooks and scripts covering everything from linear regression and MLPs up through CNNs, RNNs, GNNs, attention/Transformers, GANs, and model compression. It's aimed at Chinese-speaking students who want runnable code attached to each concept rather than a standalone library anyone would import into a project.
The pedagogical arc is genuinely wide — it goes from softmax regression to ViT/DETR to knowledge distillation under one consistent structure, which is more coverage than most CV courses attempt. Every chapter has a working notebook with Colab/Binder links, so you're not just reading theory. The shared L0CV helper package (torch.py/tensorflow.py/utils.py) cuts down on repeated boilerplate across chapters.
Almost everything — docs, comments, chapter text — is in Chinese, which is a hard wall for most of the rest of the world. The repo commits things it shouldn't: zipped datasets, hundreds of individual pikachu training PNGs, and pretrained .pth weights are all checked straight into git instead of being fetched via a script or release asset, so a clone pulls down megabytes of stuff you don't need. A lot of the chapter links point out to the author's CSDN blog posts and WeChat articles rather than keeping content self-contained in the repo, so the 'book' isn't actually complete without chasing external, rot-prone links. Last push was May 2024 — over a year stale — with no signs of the author updating it for anything post-2023 architectures.