// the find
Ewenwan/MVision
机器人视觉 移动机器人 VS-SLAM ORB-SLAM2 深度学习目标检测 yolov3 行为检测 opencv PCL 机器学习 无人驾驶
A sprawling collection of notes, READMEs, code snippets, and links covering robotics, SLAM, object detection, and autonomous driving — primarily in Chinese. It's a personal study notebook, not a usable library. Audience is Chinese-speaking CV/robotics students who want a curated reading list with occasional code examples.
The breadth of coverage is genuinely useful as a starting point: SLAM variants (ORB-SLAM2, LSD, SVO, DSO), detection architectures (YOLO v1/v2/v3, SSD, Faster-RCNN), model compression techniques (pruning, quantization, binary networks), and mobile inference (ncnn, MNN, SNPE) are all represented. The embedded interview question lists for deep learning, C++, and SLAM are surprisingly practical — they read like real hiring rubrics, not textbook summaries. The ncnn NEON/ARM optimization examples contain actual C++ and assembly-level code that is hard to find collected in one place.
This is not a codebase — most directories contain only a readme.md with links to external papers or third-party repos, so nearly everything of value is a pointer, not content. The last meaningful commit was mid-2024 and many linked resources (CSDN blogs, Baidu Pan files, academic lab pages) are already dead or paywalled. There is no build system, no tests, no consistent structure, and no way to run or verify anything here. Anyone treating this as a starting point for a real project will spend most of their time chasing broken links.