[Revised Edition] A Complete Guide to Deep Learning Computer Vision
This course will help you become a deep learning-based computer vision expert who is needed in the field by providing a deep theoretical explanation of object detection and segmentation and practical examples that can be used immediately in the field.
Learn how to use representative implementation packages for object detection and segmentation such as MMDetection, Ultralytics Yolo, and AutoML EfficientDet.
Perform image/video object detection/segmentation using OpenCV and Tensorflow Hub
Learn various difficult practical examples to be able to directly apply Object Detection/Segmentation in practice
Acquire various background knowledge that constitutes Object Detection/Segmentation
Train a Custom Dataset and Build Your Own Model Using a Variety of Implementation Packages
Experience the advantages and disadvantages of various Object Detection/Segmentation models through practical examples
Handling major data sets such as Pascal VOC and MS-COCO and converting them to TFRecord
Apply Annotation to a Dataset and Create Your Own Training Data Using the CVAT Tool
Lower the hurdles, go deeper at your core! Become a Deep Learning CNN expert.
Meet the latest revised edition Deep Learning Computer Vision Training.
Average rating 4.9★ Chosen by 1,300+ students, Inflearn Bestseller Complete Renewal in 2021!
Hello, this is Kwon Chul-min. Thanks to the support of many people, we have now released a revised edition of 'Deep Learning Computer Vision Complete Guide' . About 90% of the videos in the existing lectures have been newly created, and we will introduce improved and additional content.
Based on the feedback you have sent to the lectures, we have created a revised edition focusing on the following points.
A more detailed theoretical explanation of the topics that students have frequently asked questions about
Hands-on training based on the latest/highest performing Object Detection/Segmentation package
Reflecting the latest trends in Object Detection/Segmentation
Writing more flexible, diverse, and scalable hands-on code + more detailed explanations
Various other additional classes
The revised lecture is definitely better and more detailed than the first lecture. It will guide you into the latest deep learning-based Object Detection and Segmentation areas.
Lecture Introduction 📝
The center of deep learning computer vision technology is rapidly shifting to Object Detection and Segmentation .
▲Intelligent image information recognition ▲AI vision inspection smart factory ▲Automatic medical image diagnosis ▲Robotics ▲Autonomous vehicles, etc. Deep learning-based Object Detection and Segmentation technologies are spreading in many fields. Accordingly, leading domestic and international AI companies are also sparing no investment in this field and seeking to secure development personnel.
Two trends finally met, Object Detection & Segmentation
In recent years, the field of Object Detection and Segmentation has been developing rapidly, and the demand for talents with relevant practical skills is increasing. However, as it is a cutting-edge field that applies deep learning, there is a lack of books, materials, and lectures for learning, making it difficult to train appropriate personnel.
As a deep learning computer vision expert We will guide you so that you can be reborn.
This course consists of in-depth theoretical explanations of Object Detection and Segmentation and many practical examples that can be used immediately in the field , and will help you become a deep learning-based computer vision expert needed in the field.
Starting with an easy explanation of the concept Even to in-depth theory.
We will clearly explain everything from easy concepts to in-depth theories about the vast field of Object Detection/Segmentation, including RCNN series, SSD, YOLO, RetinaNet, EfficientDet, and Mask RCCN.
You can thoroughly learn the concepts with detailed lecture notes.
Through practical examples Maximize your deep learning practical capabilities.
There is no better way to improve your practical skills than by coding and implementing things yourself. This course consists of many practical examples that will help you maximize your practical implementation skills in Object Detection and Segmentation.
How does deep learning CNN work? Can it be applied in practice? Anyone who was worried
Deep learning based Computer Vision Solutions For those who want to develop
Deep learning image classification capabilities Up to date with the latest CV technology Those who want to expand
Going to graduate school for artificial intelligence, Deep Learning-Based CV Field Job seekers/career changers
Please check your player knowledge.
Experience with Python programming and basic understanding of deep learning CNNs are required.
Also, some experience with TF.Keras or Pytorch would be a plus.
Hard to see anywhere The latest CV technologies all in one place.
Very good performance State-of-the-art Object Detection/Segmentation implementation Practice using packages
MMDetection, Ultralytics Yolo, AutoML EfficientDet, etc.
Inference Practice Using General-Purpose OpenCV DNN and Tensorflow Hub
For various images and videos Object Detection/Segmentation Practice
Various cases where computer vision technology is actually used
With multiple custom data sets Model Training Practice
Various custom data sets
As a deep learning computer vision expert, you should be able to train models with multiple custom data sets to produce your own Object Detection/Segmentation model. You should also be able to improve and evaluate the performance of this model.
This course will teach you the ability to train custom data sets and create optimal inference models using various implementation packages.
With a self-created training data set Custom Model Training / Inference Practice
Practice with your own training data set
Using the annotation tool CVAT, we will create a training dataset that applies bounding box annotations to general images, and practice Custom Model Training and Inference using the dataset created in this way.
Practice Environment 🧰
All practical code used in the lecture was written based on the Google Colab environment.
Google Colab, Kaggle logo
We will conduct hands-on training based on GPU, and if Colab's free GPU allocation is not sufficient, we also recommend using Colab Pro. (※ Colab Pro costs about $10 per month.)
If you do not have enough Colab GPU free kernel resources, you can use the Kaggle kernel. We also provide separate practice codes made for Kaggle. You can hear more detailed information about the practice environment by referring to the Section 0 - [Setting up the practice environment] class.
Please check before taking the class!
If you do not use GPU kernels such as Colab or Kaggle, you may have difficulty following the examples. We ask for your understanding in advance.
Practice Code and Lecture Materials 👨💻
The practice code can be downloaded from https://github.com/chulminkw/DLCV_New .Reviewing the practice code in advance will help you get a sense of the level of programming required to understand the practice.
Provides 320 pages of lecture PDF textbook
The textbook used in the lecture (320 pages) can be downloaded from Lecture Section 0: Lecture Textbook .
To learn the theory There is no better way than practice.
Don't wait until you fully understand the theory of deep learning. There is no better way to learn the theory than through practice.
Once we start coding, our brains follow to understand the material. Let's implement the various practical examples presented in the lecture with me. If you listen to the lecture and implement it yourself by pressing the keyboard, the parts that felt like clouds in the sky will gradually become real.
To become an expert, sometimes (I think most of the time) you have to run before you can walk. This course will be your best companion to help you develop your career and capabilities in the field of deep learning-based computer vision.
thank you
― What Tony Stark said to Jarvis during the Iron Man suit test in <Iron Man 1>
Nó hay đến mức nực cười... Tôi đã rất tức giận với Infron khi nghe Coco Pytorch ở khu phố bên cạnh. Tôi tìm thấy niềm vui sau khi tham gia khóa học này. Cảm ơn bạn rất nhiều.
Đây là bài giảng rất hay, có thể áp dụng ngay vào thực tế. Nếu người hướng dẫn có thời gian, tôi hy vọng sẽ mở một bài giảng đề cập đến dữ liệu dựa trên RNN như giọng nói và văn bản.