Deep Learning with PyTorch – Eli Stevens, Luca Antiga, Thomas Viehmann – 1st Edition

As kids in the 1980s, taking our first steps on our Commodore VIC 20 (Eli), Sinclair Spectrum 48K (Luca) and Commodore C16 (Thomas), we saw the dawn of personal computers, learned to code and writing algorithms on ever-faster machines, and I often dreamed of where computers would take us. We were also painfully aware of the gap between what computers did in movies and what they could do in real life, and we collectively put the Rolling our eyes when the main character in a spy movie said, “Computer, get better.” Later in our professional lives, two of us, Eli and Luca, independently challenged ourselves with computer analysis. medical imaging, facing the same kind of struggle in writing algorithms that could handle the natural variability of the human body.

There were a lot of heuristics involved in choosing the best combination of algorithms that could make things work and save the day. Thomas studied neural networks and pattern recognition at the turn of the century, but later earned a PhD in mathematics doing modeling. When deep learning emerged in the early 2010s, making its initial appearance in computer vision, it began to be applied to medical image analysis tasks such as identifying structures or lesions in medical images.

It was at that time, in the first half of the decade, that deep learning appeared on our individual radars. It took me a while to realize that deep learning represented a whole new way of writing software: a new class of multipurpose algorithms that could learn to solve complicated tasks through observing data.