Language models managed to woo everyone in 2023. Moving forward, the models will hopefully start looking at the world to understand it better. Making this possible would be computer vision – a technique that can be used in various fields effectively from trying to make cars drive autonomously to detecting cancer.
Through computer vision, machines can derive information from visual inputs and then act or recommend on that. You can get started and learn advanced computer vision through several courses and resource materials, but most of them can be expensive. Here are 6 free beginners, intermediate and advanced courses on computer vision:
Computer Vision Essentials
The Great Learning course is about image processing knowledge and getting hands-on with the OpenCV library using Python for AI and machine learning. Forget the boring theory, the course provides real-world action – sampling data, messing with images, and learning the ropes of computer vision.
Instead of stopping at the basics, the course tackles everything from spotting things in pictures to figuring out what’s in computer vision. The last module will be an in-depth discussion of transfer learning.
Introduction to Computer Vision and Image Processing
IBM is offering a beginner-level course on computer vision instructed by Aije Egwaikhide and Joseph Santarcangelo. The course covers various topics such as computer vision applications across different industries, image processing and analysis techniques, Python, Pillow, and OpenCV for basic image processing, image classification, and object detection.
The course also teaches supervised learning techniques to create an image classifier. Aije Egwaikhide and Joseph Santarcangelo will collectively instruct the subject matter programme.
Advanced Computer Vision with TensorFlow
The course is designed for software and machine learning engineers to learn advanced TensorFlow features. The course covers image classification, image segmentation, object localisation, and object detection. Learners will apply transfer learning to object localisation and detection, customise existing models, and build their models to detect, localise, and label their rubber duck images.
They will also implement image segmentation using variations of the fully convolutional network (FCN), including U-Net and Mask-RCNN, to identify and detect numbers, pets, zombies, and more. Experts instruct the course in the field, which is designed for early and mid-career engineers with a basic understanding of TensorFlow.
Computer Vision with Embedded Machine Learning
The intermediate-level course, offered by a partnership among Edge Impulse, OpenMV, Seeed Studio, and the TinyML Foundation, teaches about neural networks, focusing on the classification of images and object detection in images and videos. One would also learn to deploy models to embedded systems, a field known as embedded machine learning or TinyML.
Throughout the course, students will understand how convolutional neural networks (CNNs) function and how to utilise them for image classification and object detection. The hands-on projects will provide valuable experience training custom CNNs, besides deploying them to microcontrollers and single-board computers.
Self-Driving Cars Specialisation
With over 70,000 learners already enrolled, the course offers an understanding of the engineering practices employed in the self-driving car industry. Through hands-on projects using the open-source simulator CARLA, participants will engage with real data sets from an autonomous vehicle (AV).
Throughout the program, students will get insights shared by experts from Oxbotica and Zoox. The course provides a realistic driving environment featuring 3D pedestrian modelling and various environmental conditions. Upon completion, participants will be equipped to develop their self-driving software stack.
Note that specific hardware and software specifications are necessary to effectively run the CARLA simulator, including Windows 7 64-bit (or later) or Ubuntu 16.04 (or later), a quad-core Intel or AMD processor (2.5 GHz or faster), NVIDIA GeForce 470 GTX or AMD Radeon 6870 HD series card or higher, 8 GB RAM, and OpenGL 3 or greater (for Linux computers).
Computer Vision with OpenCV Python
The OpenCV for Beginners course offers an experiential approach to computer vision, focusing on object tracking, augmented reality, face detection, optical flow, and human pose estimation. Unlike many other courses, this program is tailored to be more intuitive, making it accessible to beginners.
Upon completion, participants will get a digital certificate from OpenCV.org. The course is part of the OpenCV University, and equips participants with the foundational knowledge needed to pursue further studies in computer vision, deep learning, and AI.