AI is obviously a very popular new technology direction in recent years. Since everyone recognized the capabilities of AI a few years ago, the industry has been widely discussing how to implement AI.
We can predict that AI will play an important role in many fields and industries in the future. At present, there are some applications in business, and only search recommendation and computer vision can generate income. Therefore, there is a large manpower gap in these two directions. Due to work needs, I spent about 4 months learning machine learning and neural network related knowledge at the beginning of the year. I learned about 4-6 hours a day on weekdays and 10 hours a day on weekends. If you want to get started quickly, from your own experience, you can ignore advanced mathematics and linear algebra, because there is not much relevant knowledge involved in machine learning and deep learning. The knowledge part of vision is recommended to be divided into two parts, the first part is traditional image processing, and the second part is image processing based on deep learning. However, I found that in fact, almost 80% of CV practitioners did not learn image processing knowledge from beginning to end. Now with deep learning, there is no need to manually extract features, so many people no longer pay attention to the underlying information of the image, but directly go beyond this foundation to build a model. I think this is a misunderstanding. Images in different fields, such as OCT, MR, remote sensing, natural images, etc., have huge feature differences. I don’t understand these feature differences. How to improve and improve the accuracy after building the model? How to improve the accuracy of the original model? What about making some changes on the basis? Therefore, I think it is very important to learn the knowledge of image preprocessing and postprocessing for CV, such as image denoising, segmentation, enhancement, augmentation and so on.
Learning mentality: early is an advantage, early learning and early benefit!
However, many tutorials on the Internet are relatively fragmented. In view of this, organize a learning route and follow this route to reorganize your study plan. I believe that the level of computer vision will definitely improve qualitatively.
Chapter 1: Machine Learning and Computer Vision
Introduction to Computer Vision
technical background
- Understand the direction and hotspots of artificial intelligence
Introduction to Computer Vision
- cv profile
- cv skill tree construction
- Application field
Mathematical Foundations of Machine Learning
- Linear and Nonlinear Transformations
- Basics of Probability
- entropy
- kl divergence
- Gradient Descent
Computer Vision and Machine Learning Fundamentals
images and videos
- Image Sampling and Quantization
- filtering
- histogram
- upsampling
- downsampling
- convolution
- Histogram Equalization Algorithm
- nearest neighbor difference
- Single/Bilinear Difference
Feature Selection and Feature Extraction
- feature selection method
- filter, etc.
- Feature extraction methods: PCA, LDA, SVD, etc.
edge extraction
- Canny
- Roberts
- Sobel
- Prewitt
- Hessian features
- Haar features
camera model
- pinhole imaging model
- camera model
- lens distortion
- perspective transformation
Advanced Computer Vision and Machine Learning
Clustering Algorithm
- kmeans
- hierarchical clustering
- density clustering
- spectral clustering
Coordinate transformation and visual measurement
- Left and right handed coordinate system and transformation
- universal lock
- rotation matrix
- Quaternion
3D computer vision
- stereo vision
- multi-view geometry
- SIFT algorithm
3D Computer Vision and Point Cloud Models
- PCL point cloud model
- spin image
- 3D reconstruction
- SFM algorithm
image filter
- pass filter
- voxel filtering
- bilateral filter
- conditional filter
- radius filter
- Image Noise Addition and Noise Reduction
Detailed explanation of OpenCV
OpenCV algorithm analysis
- linear fit
- least square method
- RANSAC algorithm
- hash algorithm
- DCT algorithm
- Hamming distance
- image similarity
Chapter 2: Deep Learning and Computer Vision
Neural Networks
Deep Learning and Neural Networks
Introduction to Deep Learning
- Basic Deep Learning Architecture
- Neurons
- Detailed activation function (sigmoid, tanh, relu, etc.)
- perceptual knowledge hidden layer
- How to define the network layer
- loss function
inference and training
- Inference and Training of Neural Networks
- Detailed bp algorithm
- Normalized
- Detailed Batch Normalization
- Solve overfitting
- dropout
- softmax
- Hand pushing the training process of the neural network
Train a neural network from scratch
- Use python to implement neural network training from scratch
- Summary of experience in building neural networks
Deep Learning Open Source Framework
- pytorch
- tensorflow
- caffe
- mxnet
- hard
- Detailed explanation of optimizer (GD, SGD, RMSprop, etc.
Finally, I will share with you some artificial intelligence learning materials I have compiled for free. It has been compiled for a long time and is very comprehensive. Including some artificial intelligence basic introductory videos + AI common framework practical videos, image recognition, OpenCV, NLP, YOLO, machine learning, pytorch, computer vision, deep learning and neural network and other videos, courseware source code, well-known domestic and foreign elite resources, AI popular Papers, etc.
The following are some screenshots, click on the business card at the end of the article to follow my official account [AI Technology Planet] and send the code 321 to receive
Table of contents
1. AI Free Video Courses and Projects
2. Artificial intelligence must-read books
3. Collection of Papers on Artificial Intelligence
4. Machine Learning + Computer Vision Basic Algorithm Tutorial
Five, deep learning machine learning cheat sheet (a total of 26)
To learn artificial intelligence well, you need to read more books, do more hands-on work, and practice more. If you want to improve your level, you must learn to calm down and learn systematically slowly, so that you can gain something in the end.