One photo contest flipped the whole game — 2012: deep learning comes back
What ended the winter wasn't a new idea — it was the day data, computing and depth came together.
1 · Think first
In the 2012 ImageNet photo-recognition contest, a neural network from a University of Toronto team won with a top-5 error of 15.3%. Second place was 26.2%, so the gap was more than 10 percentage points — a gap that big was a first for that contest.
Almost nothing was newly invented. Millions of photos · gaming graphics cards · a deeply stacked neural network — that year, those three came together for the first time.
After that, research on images switched over to neural networks across the board. It was the day the second winter from yesterday truly ended.
2 · Experiment · Before & after
Try it — before and after 2012
It's just like watching a before-and-after video of a game patch. Pick a topic, then switch between “Up to 2011” and “2012.” You'll see right away what changed.
Figure
In one picture
The error rate had been shrinking a little each year — then in 2012, one entry suddenly plunged.
3 · Quiz
Quick check
Three questions. Pick an answer and it's marked right away, with the reason why.
Today in one line
Today in one line
What ended the winter wasn't a new idea —
it was the day data, computing and depth came together.
Learn more
- The official 2012 contest results. The winning team SuperVision's top-5 error of 15.3% is listed here — ILSVRC 2012 Results
- The paper describing that network. It's where you'll read that it was trained for 5–6 days on two graphics cards — ImageNet Classification with Deep Convolutional Neural Networks (NIPS 2012)
- The paper introducing ImageNet itself, the collection of millions of photos — ImageNet: A Large-Scale Hierarchical Image Database (2009)