20210605 What does it mean to actually see something? Since at least the 1980’s we have been getting our computers more competent at collecting an image through a digitally connected camera and processing the image in some way. It has been pretty straight forward to identify colour in images and generally identify boundaries within the image. Through complex algorithmic approaches we were able to basically identify objects within the image, identifying, chairs, and balls and so in.
With the advent of deep neural networks, that is networks of neurons with very many layers and nodes, we’ve replaced the algorithmic approaches (which us humans writing the algorithms) with automated feature learning and algorithm generating approaches that we little understand. Yet, building these models from massive collections of data utilising considerable amounts of compute, the models can now often exhibit better than human performance.
In the future we might also hope that we will discover how this is achieved and so gain actual knowledgeable insights into intelligence.
In this chapter we explore the current state-of-the-art in computer vision with the aim to explain it all simply. We will often highlight how you can explore the technology yourself through the MLHub repository. Reference will frequently be made to the companion sections of the Machine Learning Hub Desktop Survival Guide sections on Computer Vision. Lear about computer vision by doing it, and find a few useful tools that might actually become part of your daily toolkit.
Your donation will support ongoing availability and give you access to the PDF version of this book. Desktop Survival Guides include Data Science, GNU/Linux, and MLHub. Books available on Amazon include Data Mining with Rattle and Essentials of Data Science. Popular open source software includes rattle, wajig, and mlhub. Hosted by Togaware, a pioneer of free and open source software since 1984. Copyright © 1995-2022 Graham.Williams@togaware.com Creative Commons Attribution-ShareAlike 4.0