AlexNet wins ImageNet
- neural networks
- vision
Alex Krizhevsky, Ilya Sutskever and Geoffrey Hinton developed the deep convolutional network commonly called AlexNet, whose 2012 ImageNet competition result sharply improved large-scale image classification. Training combined a substantial labeled dataset, GPU computation and architectural and optimization choices that made a large neural model practical. Its result demonstrated the power of learning visual representations rather than relying entirely on hand-designed features. It helped redirect computer-vision research toward deep learning, while building on decades of earlier neural-network and convolutional work. This is a benchmark and research milestone, not the origin of all artificial intelligence. [Original paper](https://papers.nips.cc/paper_files/paper/2012/hash/c399862d3b9d6b76c8436e924a68c45b-Abstract.html). *Image: Wikimedia Commons (AlexNet's architecture, from Zhang, Lipton, Li and Smola, Dive into Deep Learning; CC BY-SA 4.0).* **Timeline anchor:** 30 September 2012 CE. Approximate day: the 2012 results. **References:** [Source 1](https://en.wikipedia.org/wiki/AlexNet).
- Date precision
- Approximate day: the 2012 results.
Sources
Related
On other timelines
- AlexNet on The history of computer science by @athan
- AlexNet demonstrates GPU-trained large-scale image classification on GPU and ASIC programming · parallel hardware, standards and applications by @athan