Machine Learning, Computer Vision and Artificial Intelligence

• Introductions

  • Artificial Intelligence (AI), Machine Learning (ML) and Deep Learning (DL)
  • Artificial Neural Network
  • Deep Neural Network
  • Generative Model
    • Generative Adversarial Network-GAN
    • Encoder-Decoder, Autoencoder and U-Net
    • Diffusion Models
    • Stable Diffusion
    • ChatGPT
  • Machine Learning Performance Evaluation Metrics
    • Evaluation for Classification
    • Evaluation for Regression
    • Mean Average Precision (mAP) for Image Recognition
  • Machine Learning Distance Metrics
  • Overfitting and Underfitting
    • Bias and Variance
    • Train Test Split
    • Cross Validation
    • Feature Selection and Dimensionality Reduction (ML)
    • Normalization (DL)
    • Regularization
      • Ridge Regression and Lasso Regression
    • Dropout for Neural Networks (DL)
    • Data Augmentation (DL)
  • Gradient Descent
    • Backpropagation
    • Cost Function and Loss Function
      • Information, Entropy, Cross Entropy, Categorical CE, Binary CE
      • KL Divergence and JS Divergence
      • Focal Loss
    • Learning Rate
    • Batch Gradient Descent (BGD)
    • Stochastic Gradient Descent (SGD)
    • MiniBatch Gradient Descent
    • Convex Function
    • Local Minima and Saddle Point
    • Vanishing and Exploding Gradient
    • Exploding Gradient
    • Optimization Algorithms SGD, Adam, RMSProp

• LLM

  • Introductions and Tutorials
    • Attention
      • FlashAttention
      • PagedAttention
      • Grouped Query Attension
      • Visualizing A Neural Machine Translation Model (Mechanics of Seq2seq Models With Attention)
      • The Attention Mechanism from Scratch
    • Key-Value Cache
    • AWQ: Activation Aware Quantization
    • RMS Normalization
    • RoPE: Rotary Position Embedding
    • Token Embedding
      • Word Embedding
        • Word2Vec
        • What Are Word Embeddings for Text?
        • The Illustrated Word2vec
    • Transformer
      • Vision Transformer
    • Knowledge Distillation
  • HuggingFace
  • DeepSeek
  • LLM Inference Optimizations
    • C/C++ based
      • llama.cpp
        • End-to-End flowchart
        • llama_decode deep dive
        • build Qwen graph
      • llamafile
      • PowerInfer
    • Python based
      • vLLM
      • SGLang
      • KTransformers
      • lmdeploy
    • Rust based
      • LlamaEdge
    • Kernel Optimization
      • Marlin: a Mixed Auto-Regressive Linear kernel, an extremely optimized FP16xINT4 matmul kernel
  • ax-llm

• VLM

  • Vision Language Models Explained
  • CLIP: Contrastive Language-Image Pre-Training
    • SigLIP: Sigmoid Loss for Language Image Pre-Training
    • BLIP and BLIP-2: Boostrapping Lanugage-Image Pre-training
  • BAAI: Aquila-VL
  • DriveVLM: The Convergence of Autonomous Driving and Large Vision-Language Models
  • Senna: Bridging Large Vision-Language Models and End-to-End Autonomous Driving
  • LLaVA: Large Language and Vision Assistant
  • LLaVA-OneVision: Easy Visual Task Transfer

• Conversational AI

  • Introduction

• Machine Learning

  • IRIS classification with Scikit-learn quickstart
    • IRIS with Different Machine Learning Algorithms
  • Data Loading
  • Data Exploration
    • Understanding Data with Statistics: Summary, Distributions, Correlations and Skewness
    • Understanding Data with Visualizaiton: Histogram, Density, Box and Correlation Matrix Plots
  • Data Preprocessing
    • Data Cleaning
      • Data Cleaning: Outlier Detection and Removal
    • Data Transformation
      • Scaling in Data Preprocessing
      • Discretization in Data Preprocessing
      • Categorical Data Encoding
        • ColumnTransformer and Pipeline for Mixed Data Types
    • Data Reduction
      • Principal Component Analysis (PCA)
        • PCA with Scikit-learn on IRIS dataset
        • PCA: feature transformation intuitive guide
        • Statistical and Mathematical Concepts behind PCA
        • Image Compression Using PCA
      • Linear Discriminant Analysis (LDA)
        • LDA with scikit-learn on Wine dataset
        • Statistical and Mathematical concepts behind LDA
  • Hyper-Parameters Tuning
    • Bayesian Optimization of Hyperparameter Tuning With scikit-opitmize
    • Bayesian Optimization of Hyperparameter Tuning With Hyperopt
  • Machine Learning Models
    • Linear Regression
    • Logistic Regression
    • Perceptron
    • Support Vector Machines
    • Decision Tree
      • Decision Tree with IRIS dataset
    • Random Forest
    • XGBoost
    • K-Nearest Neighbors (KNN)
    • Naive Bayes Classification
    • Clustering
      • Centroid Model: K-means
      • Mean Shift
      • Connectivity Model: Hierarchical Clustering
      • Density Model: DBSCAN
    • Association Rules
    • Ensemble
      • Bagging
        • Random Forest
      • Boosting
        • XGBoost
      • Voting
  • Machine Learning Model Explainability
    • Machine Learning Pipeline with Explainability
    • SHAP: Explain Any Machine Learning Model
  • Concepts
    • Information, Entropy, Cross Entropy, Categorical CE, Binary CE
    • KL Divergence and JS Divergence
    • Vector Norm
    • Correlation
    • Focal Loss
    • Synthetic Minority Oversampling Technique (SMOTE)
    • Zero-Shot Learning

• Deep Learning

  • A Visual and Interactive Guide to the Basics of Neural Networks
  • A Visual And Interactive Look at Basic Neural Network Math
  • Batch Size, Training Steps and Epochs
  • Hyperparameters
    • Deep Learning Tuning Playbook
  • Activation Functions
  • Convolution
    • 1D Convolution
    • Convolution Filters
  • Pooling
  • Normalization
  • Residual Block and Inverted Residual Block
  • Deep Learning Framework
    • PyTorch vs Tensorflow
    • PyTorch
      • PyTorch Installation
      • PyTorch Quickstart
      • PyTorch Computational Graph
        • Update the parameters of a neural netowrk
        • Leaf Tensor in-place update
      • PyTorch Access A Layer by the Module Name
        • modules() vs children()
        • nn.Parameter
      • PyTorch Extract an Intermediate Layer
      • nn.Module vs nn.Functional
      • ModuleList and ParameterList
    • Tensorflow
      • Tensorflow model optimization for inference
      • logits, softmax and softmax_cross_entropy_with_logits
      • Protobuf and Flat Buffers
      • tf.placeholder vs tf.Variable
      • Graph vs GraphDef
      • Save and Restore Tensorflow Models
      • TFRecord to Store and Extract Data
      • Convert a Global Average Pooling layer to Conv2D
    • ONNX
      • Create a toy model with LayerNormalization
      • Get intermediary layer output of an ONNX model
      • NNEF and ONNX: Similarities and Differences
  • Quantization
    • Quantization Arithmetic
    • PyTorch
      • Selective Quantization
      • Introduction to Quantization on PyTorch
      • Practical Quantization in PyTorch
      • PyTorch Numeric Suite Tutorial
      • Torch Quantization Design Proposal
      • Eager Mode Quantization
        • MobileNetV2 QAT on CIFAR-10
        • Resnet18 QAT on CIFAR-10
        • Static quantization with Eager Mode in PyTorch
      • FX Graph Graph Mode Quantization
    • ONNX
      • ONNX Runtime Qunatization Example MobilenetV2 with QDQ Debugging
      • Mobilenet v2 Quantization with ONNX Runtime on CPU
    • TensorRT
      • Docker Environment Setup
      • VGG16 QAT on CIFAR-10
      • YOLOv5 Quant Example
  • Stable Diffusion
    • The Illustrated Stable Diffusion
  • RepVGG: Making VGG-style ConvNets Great Again
  • Compiler
    • GLOW

• Computer Vision

  • Overview

  • Object Detection

    • Concepts
      • Anchor Boxes
      • Feature Pyramid Network (FPN)
      • Focal Loss
      • Intersection over Union (IoU)
        • IoU sample notebook
      • Mean Average Precision (mAP)
      • Non Maximum Suppression (NMS)
        • NMS in PyTorch
      • Region Proposal Network (RPN)
    • YOLOv7
      • YOLOv7 QAT
    • YOLOv5
      • Object Detection using YOLOv5 and OpenCV DNN
      • YOLOv5 Post Training Quantization with ONNX
      • YOLOv5 QAT
    • SSD - Single Shot Detector
    • RetinaNet
      • Face mask detection with RetinaNet example

  • Classification

    • Transfer Learning with PyTorch tutorial

  • Segmentation

    • Color Segmentation
      • Color Image Segmentation in HSV Color Space
    • Deep Adversarial Training for Multi-Organ Nuclei Segmentation in Histopathology Images

  • Image Translation

    • Pix2Pix - Image-to-Image Translation with Conditional Adversarial Networks
    • CycleGAN - Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks

  • 3D

    • 3D Technology Map
    • Point Cloud
      • Introduction to Point Cloud
      • Estimate Point Clouds From Depth Images
    • PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation
    • PointPillars
      • PointPillars Introduction
    • Depth Estimation from Stereo Images

  • Deepface

  • Data Annotation

    • Computer Vision Annotation Tool (CVAT)
    • LabelImg
    • Autonomous driving
      • 2D Bounding Box
      • Lane Line
      • Semantic Segmentation
      • Video Tracking Annotation
      • 3D Point
      • 3D Object Recognition (3D Cube)
      • 3D Segmentation
      • Sensor Fusion: Cuboids/Segmentation/Tracking

• Image Processing

  • Color Temperature Kelvin to RGB
  • HSV Color Space
  • Color Correction and Calibration
    • Histogram Matching
    • Color Correction with Color Card
  • Focus
    • Image Focus Checking
  • Utilities
    • Split images
  • Color Science with Python

• Text and NLP

  • TF-IDF: Term Frequency, Inverse Document Frequency
  • Doc2Vec
  • The Illustrated BERT, ELMo, and co. (How NLP Cracked Transfer Learning)
  • How GPT-3 Works - Visualizations and Animations
  • The Illustrated GPT-2 (Visualizing Transformer Language Models)

• Data Structure and Algorithms

  • Heap and Heap Sort in Python

• Linear Algebra

  • Matrix Factorization and SVD

• Data Mining

  • Association Rules

• Kalman Filter

  • Kalman Filter Tutorial
  • Kalman Filter Implementation in Python

Miscellaneous

Resources

• Netron
• Hugging Face
• PyTorch Image Models
• Learn OpenCV : C++ and Python Examples
• Machine Learning Tutorial

Dataset

• Kaggle Datasets
• UCI Machine Learning Repository
• Registry of Open Data on AWS
• Google's Dataset Search Engine
• Microsoft Research Open Data
• Awesome Public Datasets
• scikit-learn dataset
• Computer Vision
  • VisualData Discovery
  • COCO Common Objects in Context
  • ImageNet
  • PASCAL VOC
  • YouTube-Objects
  • Face
    • IMDB-WIKI – 500k+ face images with age and gender labels
    • WIDER FACE: A Face Detection Benchmark
    • FDDB: Face Detection Data Set and Benchmark
  • Super-Resolution
    • DIV2K dataset: DIVerse 2K resolution high quality images for super-resolution
  • Object Segmentation
    • Cityscapes
    • KITTI
    • ADE20K
    • A Dataset and a Technique for Generalized Nuclear Segmentation for Computational Pathology
  • Roboflow public datasets

Online books

• Dive into Deep Learning
• Neural Networks and Deep Learning by Michael Nielsen
• Deep Learning by Ian Goodfellow, Yoshua Bengio and Aron Courville
• Computer Vision: Algorithms and Applications, 1st ed. by Richard Szeliski
• Digital Image Processing by University of Tartu

Online Courses

• Machine Learning Crash Course by Google
• CS231n: Convolutional Neural Networks for Visual Recognition by Stanford
  • Programming Assignments and Lectures for Stanford's CS 231: Convolutional Neural Networks for Visual Recognition by Standford
• Deep Learning Nanodegree Foundation Program by Udacity
• Deep Learning course: lecture slides and lab notebooks by University Paris Saclay