Computer vision - free course from Otus, training 4 months, Date: December 5, 2023.

During the course, you will train neural networks to solve problems:
- classification and segmentation of images
- detection of objects in images
- tracking objects on video
- processing of three-dimensional scenes
- generating images and attacks on trained neural network models
You will also learn how to use the main frameworks for creating neural networks: PyTorch, TensorFlow and Keras. Map of Data Science courses at OTUS

Who is this course for?

For Machine Learning professionals who:

• Want to specialize in Computer Vision
• Already using Deep Learning practitioners and want to expand and systematize knowledge
• The course will allow you to switch from classic machine learning tasks such as credit scoring, CTR optimization, fraud detection and etc., and get into the developing field of Data Science, where all the most interesting things are happening now and new careers are opening up horizons.

The training will give you the necessary competencies to apply for jobs that require professional computer vision system development skills. In different companies, specialties are called differently, the most common options are: Deep learning engineer, Computer Vision Engineer, AI Research Engineer [Computer Vision, Machine Learning], research programmer, Deep Learning/Computer Vision.

How is the course different from others?
Preparing to solve combat missions: how to launch a neural network in the cloud and adapt the model for different platforms
In-depth knowledge and modern approaches to Computer vision technologies
Completed project work that can be added to your portfolio
Funny examples, a fountain of ideas and cyberpunk universes at your fingertips - 4 months will fly by in one breath!

During the course you:
You will work with open datasets for various Computer Vision tasks
You will understand the operating principles and options of convolutional and pooling layers, including those specific to object detection and segmentation tasks.
Learn to apply attention mechanism in convolutional networks.
Find out what ideas underlie modern convolutional networks (MobileNet, ResNet, EfficientNet, etc.)
You will understand DL approaches to object detection - study the R-CNN family, real-time detectors: YOLO, SSD. You can also implement an object detector yourself.
Learn to solve the Deep Metric Learning problem using Siamese networks. Learn what triplet loss and angular loss are.
Gain experience in solving image segmentation problems: U-Net, DeepLab.
Learn to apply fine tuning, transfer learning and collect your own datasets for object detection and Image segmentation, metric learning tasks.
You will work with generative adversarial networks. Understand how GANs can be used for adversarial attacks and how to implement super resolution GANs.
Learn to run models on the server (tensorflow serving, TFX). Get acquainted with frameworks for optimizing neural networks for inference on mobile/embedded devices: Tensorflow Lite, TensorRT.
Explore architectures for defining Facial Landmarks: Cascade shape regression, Deep Alignment Network, Stacked Hourglass Network

Anna Zakutnyaya

1

well

She graduated from the Master's program in quantitative finance at the National Research University Higher School of Economics. Since university he has been interested in machine learning and deep learning problems. Managed to work on various projects: developed a pipeline for detection and recognition of paintings; integrated recognition module...

Alexander Sizov

3

course

Experienced developer, scientist and Machine/Deep learning expert with experience in recommender systems. He has more than 30 scientific publications in Russian and foreign languages, defended his PhD thesis on the topic of analysis and...

Anton Vitvitsky

2

course

Expert in computer vision and deep learning, certified software engineer and candidate of physical and mathematical sciences. From 2012 to 2017, he worked in facial recognition at WalletOne, whose solutions were supplied to businesses in South...