深度学习/计算机视觉相关的博文/在线图书等的汇总,主要是个人比较感兴趣的环境理解/视频理解方向的
目标检测
- mmdetection系列
经典目标检测/实例分割算法原理简介和统一实现 https://zhuanlan.zhihu.com/p/337375549 - 样本不均衡:https://ranmaosong.github.io/2019/07/20/cv-imbalance-between-easy-and-hard-examples/
- 各种NMS:https://blog.csdn.net/qq_40263477/article/details/103881569
- Trick总结:https://zhuanlan.zhihu.com/p/137768226
- Anchor:https://zhuanlan.zhihu.com/p/63024247
量化
- Pytorch量化感知训练:https://leimao.github.io/blog/PyTorch-Quantization-Aware-Training/
- 神经网络量化基础:https://leimao.github.io/article/Neural-Networks-Quantization/
- Pytorch后量化:https://leimao.github.io/blog/PyTorch-Static-Quantization/
Backbone/网络结构
- ViT:https://abhaygupta.dev/blog/vision-transformer
- 经典Transtormer介绍:http://jalammar.github.io/illustrated-transformer/
- DCN:https://www.jianshu.com/p/206e7b0cb433
- 动态卷积比较:https://zhuanlan.zhihu.com/p/142196208
- GCN:https://jonathan-hui.medium.com/graph-convolutional-networks-gcn-pooling-839184205692
框架
- Pytorch内部实现:http://blog.ezyang.com/2019/05/pytorch-internals/
轻量化模型
- MobileNetV2: https://zhuanlan.zhihu.com/p/98874284
性能/加速
- Roofline介绍:https://zhuanlan.zhihu.com/p/33693725
- 卷积神经网络的复杂度分析:https://zhuanlan.zhihu.com/p/31575074
- Roofline:https://zhuanlan.zhihu.com/p/34204282
- 内存格式:https://oneapi-src.github.io/oneDNN/dev_guide_understanding_memory_formats.html
视频理解
- 视频分析方面专栏:https://www.zhihu.com/column/wzmsltw
跟踪
- 相关滤波:https://zhuanlan.zhihu.com/p/59624151
贝叶斯网络
- 变分推断:https://blog.evjang.com/2016/08/variational-bayes.html
AutoML/NAS
- NAS:https://lilianweng.github.io/lil-log/2020/08/06/neural-architecture-search.html
高质量博客
- https://evjang.com/
- https://www.computervisionblog.com/
- https://lilianweng.github.io/lil-log/archive.html
语义分割
- 谷歌视频语义分割方案:https://ai.googleblog.com/2018/03/mobile-real-time-video-segmentation.html
Online Book
- AI笔记:http://www.huaxiaozhuan.com/
- 复旦邱锡鹏:https://nndl.github.io/
- http://neuralnetworksanddeeplearning.com/
传感器/Camera等
- iPhone dTOF 数据获取、精度分析:https://www.it-jim.com/blog/iphones-12-pro-lidar-how-to-get-and-interpret-data/
课程
- CMU计算摄影学:http://graphics.cs.cmu.edu/courses/15-463/
|