The source code of "Alexnet," which is the catalyst for the development of artificial intelligence (AI) based on "deep learning," which is currently bringing about a major change in our lives, has been released. This is the original code that won the 2012 ImageNet competition.

On the 20th (local time), the Computer History Museum (CHM) in Mountain View, Silicon Valley, announced that it would release Alexnet source code as an open source in cooperation with Google. This can be found in the git-hub of CHM.

AlexNet is an image recognition software based on an artificial neural network created by Alex Kriezewski and Ilya Sutzkeber, a graduate student

studying under Professor Jeffrey Hinton at the University of Toronto in 2012.

Since 2010, Alex Net has participated in the "Image Net" competition held by Stanford University professor Feifery and beat other artificial intelligence with overwhelming performance. The two used two NVIDIA 3D graphics cards to learn the artificial neural network.

At the time, the only image recognition AI using an artificial neural network called deep learning was 'AlexNet'. However, the paradigm of AI development has completely changed to deep learning as deep learning has proven to have excellent performance in practice.

In addition, Nvidia's GPU, which was used for learning artificial neural networks in deep learning, has become a key infrastructure for researching and servicing AI, and Nvidia's sales have grown rapidly with advances in AI. Nvidia once ranked No. 1 in global corporate market capitalization.

After Alexnet's victory in 2012, Google recruited them as AI researchers in the form of acquiring DNN Research, which was founded by three researchers, and Alexnet's copyright became part of Google in the process. With Google's help, CHM released it as an open source, just like the source code used in the 2012 ImageNet competition.