This article is mainly based on my own experience, focusing on the introduction of computer vision learning resources, if you follow this route to learn, I believe that it will greatly promote the improvement of your computer vision knowledge level.

Before we dive into computer vision, let’s take a look at machine learning and python basics.

Frameworks

While you don’t have to decide which framework to choose from the start, it is necessary to apply the new knowledge in practice.

There are not many options for frameworks, mainly PyTorch or Keras (TensorFlow). Pytorch may require more code, but it has a lot of flexibility in terms of returns, so we can learn how to use Pytorch first. In addition, pyToch is also commonly used by most deep learning researchers.

Albumentation (image Enhancement libraries) and Catalyst (frameworks, the high-level API at the top of PyTorch) are also common tools in our long learning of computer vision, and we can learn and use them first, especially the first.

hardware

  • Nvidia GPU 10xx + :($300 +)
  • Kaggle kernel (free) : only 30 hours per week (www.kaggle.com/kernels)
  • Google Colab (free) : 12 hours session limit, limit the use of a week long variable (colab.research.google.com/notebooks/i…

Theory and Practice

Online courses
  • CS231n is an excellent online teaching course that covers all the necessary basics of computer vision and is an online video from YouTube. The course also includes after-class exercises, so beginners don’t have to complete the exercises first. (free)
    • cs231n.stanford.edu/
  • Ai is another course we should learn. Fast. Ai is an advanced framework on top of PyTorch, but they change their apis too frequently and lack documentation makes them inconvenient to use. However, it is a good idea to spend some time looking at the theory and useful techniques of the course. (free)
    • course.fast.ai/

In pursuing these courses, I suggest that you put theory into practice and apply it to one of these frameworks.

Articles and code
  • ArXiv.org – For all the latest information. (Free) (arxiv.org/)
  • (Paperswithcode.com/sota)
    • The latest developments in the most common deep learning tasks, not just computer vision. (free)
  • Github — For code in practice, you’ll find it here. (free) github.com/topics/comp…
books

There aren’t many books to read, but I believe both of these books will be useful, whether you choose to use PyTorch or Keras

  • Python deep learning by FrancoisChollet, Keras creator and Google AI researcher. Easy to use and may gain insights you didn’t know before. (Not free)
    • www.amazon.com/Deep-Learni…
  • Pytorch Deep Learning by PyTorch Team Eli Stevens and Luca Antiga (free)
    • Pytorch.org/deep-learni…
Kaggle
  • www.kaggle.com/competition… Kaggle is a well-known online platform for various machine learning competitions, many of them about computer vision. Even if you don’t finish the course, you can start participating in the contest, as there will be a lot of open kernels (end-to-end code) from the contest that you can run directly from the browser. (free)

Challenging learning Style (recommended)

The alternative approach may be difficult, but it allows you to gain knowledge in different areas of computer vision, and you can study specific areas of computer vision for your particular field of study. (Little blogger remindsHere is a list of classic projects in each field of vision.

Try reading and replaying the following articles, you will learn a lot. Help forward, I hope to help you.

The network architecture
  • AlexNet: cca shut. Nips. Cc/paper / 4824 -…
  • ZFNet: arxiv.org/abs/1311.29…
  • VGG16: arxiv.org/abs/1505.06…
  • ResNet: arxiv.org/abs/1704.06…
  • GoogLeNet: arxiv.org/abs/1409.48…
  • Inception: arxiv.org/abs/1512.00…
  • Xception: arxiv.org/abs/1610.02…
  • MobileNet: arxiv.org/abs/1704.04…
Semantic segmentation
  • FCN: arxiv.org/abs/1411.40…
  • SegNet: arxiv.org/abs/1511.00…
  • UNet: arxiv.org/abs/1505.04…
  • PSPNet: arxiv.org/abs/1612.01…
  • DeepLab: arxiv.org/abs/1606.00…
  • ICNet: arxiv.org/abs/1704.08…
  • ENet: arxiv.org/abs/1606.02…
Generative adversarial network
  • GAN: arxiv.org/abs/1406.26…
  • DCGAN: arxiv.org/abs/1511.06…
  • WGAN: arxiv.org/abs/1701.07…
  • Pix2Pix: arxiv.org/abs/1611.07…
  • CycleGAN: arxiv.org/abs/1703.10…
Target detection
  • RCNN: arxiv.org/abs/1311.25…
  • Fast – RCNN: arxiv.org/abs/1504.08…
  • Faster – – RCNN: arxiv.org/abs/1506.01…
  • SSD: arxiv.org/abs/1512.02…
  • YOLO: arxiv.org/abs/1506.02…
  • YOLO9000: arxiv.org/abs/1612.08…
  • Examples of segmentation
  • Mask – RCNN: arxiv.org/abs/1703.06…
  • YOLACT: arxiv.org/abs/1904.02…
Attitude estimation
  • PoseNet: arxiv.org/abs/1505.07…
  • DensePose: arxiv.org/abs/1802.00…

The original link: towardsdatascience.com/guide-to-le…

Welcome to panchuangai blog: panchuang.net/

Sklearn123.com/

Welcome to docs.panchuang.net/