What is AI Summer
AI Summer is a free educational blog with one single purpose. To help you learn everything you need to know about Deep Learning
If you want to become a Machine Learning Expert, a Data Scientist or simply stay updated on the latest trends in the field, this is the site for you.
Let’s build the most amazing AI solutions together.
Learn Deep Learning
Discover the fundamental principles behind Deep Neural Networks both from a programming and a mathematical point of view. Then you can proceed to more advance concepts and state of the art techniques
Build AI Products
Develop and deploy real-life AI solutions on a variety of fields such as Computer Vision, Natural Language Processing, Robotics and Healthcare.
Add Machine Learning to your business
Utilize AI into your products and gain invaluable insights about your business performance. Or build entirely new Deep Learning startups that will provide value to our society.
Stay updated on the field
Explore the most promising and recent Artificial Intelligence research as well as the best Machine Learning products and tools.
Explore our content
An Artificial Intelligenge hub where you can find and learn anything related to Deep Learning. From fundamental principles to state of art research and real-life applications
Attention and Transformers
Generative Learning
Machine Learning
Reinforcement Learning
Software
How Attention works in Deep Learning: understanding the attention mechanism in sequence models
New to Natural Language Processing? This is the ultimate beginner’s guide to the attention mechanism and sequence learning to get you started
How Transformers work in deep learning and NLP: an intuitive introduction
An intuitive understanding on Transformers and how they are used in Machine Translation. After analyzing all subcomponents one by one such as self-attention and positional encodings , we explain the principles behind the Encoder and Decoder and why Transformers work so well
How the Vision Transformer (ViT) works in 10 minutes: an image is worth 16x16 words
In this article you will learn how the vision transformer works for image classification problems. We distill all the important details you need to grasp along with reasons it can work very well given enough data for pretraining.
GANs in computer vision - Introduction to generative learning
The first article of the GANs in computer vision series - an introduction to generative learning, adversarial learning, gan training algorithm, conditional image generation, mode collapse, mutual information
GANs in computer vision - Conditional image synthesis and 3D object generation
The second article of the GANs in computer vision series - looking deeper in generative adversarial networks, mode collapse, conditional image synthesis, and 3D object generation, paired and unpaired image to image generation.
GANs in computer vision - semantic image synthesis and learning a generative model from a single image
The sixth article-series of GAN in computer vision - we explore semantic image synthesis and learning a generative model from a single image
In-layer normalization techniques for training very deep neural networks
How can we efficiently train very deep neural network architectures? What are the best in-layer normalization options? We gathered all you need about normalization in transformers, recurrent neural nets, convolutional neural networks.
A journey into Optimization algorithms for Deep Neural Networks
An overview of the most popular optimization algorithms for training deep neural networks. From stohastic gradient descent to Adam, AdaBelief and second-order optimization
Intuitive Explanation of Skip Connections in Deep Learning
What are skip connections, why we need them and how they are applied to architectures such as ResNet, DenseNet and UNet.
The secrets behind Reinforcement Learning
The central idea behind reinforcement learning and an overview of its algorithms
Q-targets, Double DQN and Dueling DQN
Fixed Q-targets, Double DQN, Dueling DQN, Prioritized Replay
Unravel Policy Gradients and REINFORCE
Explore Policy-based methods and dive into policy gradients
Data preprocessing for deep learning: How to build an efficient big data pipeline
How to develop high performance input pipelines in Tensorflow using the ETL pattern and functional programming
Distributed Deep Learning training: Model and Data Parallelism in Tensorflow
How to train your data in multiple GPUs or machines using distributed methods such as mirrored strategy, parameter-server and central storage.
How to use uWSGI and Nginx to serve a Deep Learning model
Serving a Tensorflow model to users with Flask, uWSGI as a web server and Nginx as a reverse proxy. Why we need both uWSGI and Flask, why we need Nginx on top of uWSGI and how everything is connected together?
Learn how Machine learning works
Start from the basics and learn about architectures such as Convolutional Networks and LSTMs.
Continue with more advance concepts like Reinfocement and Generative Learning by diving into state of the art research papers.
Finally it’s time to put your skills into practice by developing Computer Vision and Natural Language Processing applications using the most popular frameworks.
Autoencoders
Autoencoders are an unsupervised type of network that can learn compact representation of the data features. They can be deterministic or probabilistic
Attention and Transformers
Attention and Transformers have already been the standard in NLP applications and they are entering Computer Vision as well
Business
Artificial Intelligence is transforming business processes and will be a major driving force for future companies. Why stay behind?
Career
Become a Deep Learning Researcher or a Machine Learning Engineer. Everything you need to know to land your dream job.
Convolutional Neural Networks
A class of deep networks that use spatial structure and can be thought as regularized semi-connected feed forward networks. They have been extensively used in Computer vision applications.
Computer Vision
Computer Vision is the field that is dominated by Deep Learning. Face recognition, image classification, video prediction are only a tiny portion of applications
Data Processing
Developing high-performant big data pipelines using Tensorflow or Pytorch
Generative Learning
Deep Learning can be used to generate complete new real-like data such as images, text and more. But how is that possible?
Generative Adversarial Networks
GANs are constructed by two neural networks that compete against each other in an adversarial game, and are proven to be ideal for generating novel data.
Graph Neural Networks
GNNs are able to extract features from graphs and produce invaluable insights
Machine Learning
Funtamental Machine Learning principles and concepts that are extended into Deep Neural Networks
Medical
Accelerate healthcare progress by applying AI into disease prediction, treatment, medical images and electronic health records.
MLOps
Best practices on Machine Learning infrastructure. How to build, maintain and scale production-ready deep learning systems.
Natural Language Processing
Text generation, language translation, spoken language understanding, sentiment analysis and dialogue systems
Pytorch
Implement basic Deep Learning models and advanced real-life applications with Pytorch
Recurrent Neural Networks
Recurrent Neural Networks are deep networks that contain loop connections between nodes. Because of that, they can use their internal memory to process sequences of inputs.
Reinforcement Learning
A subfield of Machine Leaning that focuses on how agents interacting with the environment and how to make the best decisions depending on the end goal.
Resources
Curated collections of internal and external resources. The ultimate AI hub.
Software
Sharpen your software engineering skills and build robust models. MLOps, data processing and model deployment should definitely be in you arsenal
Tensorflow
Learn Tensorflow and Keras for building Deep Learning applications
Unsupervised Learning
A category of ML algorithms that analyze and discover patterns from unlabeled data
More articles
An overview of Unet architectures for semantic segmentation and biomedical image segmentation
Learn everything about one of the most famous convolutional neural network architectures that is widely used on image segmentation.
How Graph Neural Networks (GNN) work: introduction to graph convolutions from scratch
Start with Graph Neural Networks from zero and implement a graph convolutional layer in Pytorch
Best Resources to Learn Deep Learning Theory
A curated list of the best courses and books to learn deep learning
JAX vs Tensorflow vs Pytorch: Building a Variational Autoencoder (VAE)
A side-by-side comparison of JAX, Tensorflow and Pytorch while developing and training a Variational Autoencoder from scratch
Why multi-head self attention works: math, intuitions and 10+1 hidden insights
Learn everything there is to know about the attention mechanisms of the infamous transformer, through 10+1 hidden insights and observations
Build a Transformer in JAX from scratch: how to write and train your own models
How to develop and train a Transformer with JAX, Haiku and Optax. Learn by example how to code Deep Learning models in JAX
JAX for Machine Learning: how it works and why learn it
An introduction to JAX, its best features alongside with code snippets for you to get started
Explainable AI (XAI): A survey of recents methods, applications and frameworks
What is Explainable Artificial Intelligence (XAI), what are the most popular methods, where and how can it be applied
How Positional Embeddings work in Self-Attention (code in Pytorch)
Understand how positional embeddings emerged and how we use the inside self-attention to model highly structured data such as images