1989
LeNet convolutional networks
Yann LeCun applied convolutional neural networkA model of connected layers that learns from examples — the core of modern deep learning.s to handwritten digit recognition in 1989 — the LeNet architecture that foreshadowed AlexNet.
What it was for
LeNet ran on bank check readers for years before GPUs made deep CNNs mainstream. LeCun's work at AT&T Bell Labs proved convnets worked in production — not just theory.
People
- Yann LeCun — researcher
Why it's here
LeNet was the first widely deployed convolutional neural networkA model of connected layers that learns from examples — the core of modern deep learning..
Why it mattered
It established CNNs as the architecture for spatial pattern recognition.
What it solved
Handwritten character recognition needed shift-invariant feature learning.
Media
ImageYann LeCunJérémy Barande, CC BY-SA 2.0, via Wikimedia Commons