On April 19, Facebook officially announced the open source Caffe2, its new deep learning framework, at the F8 Developer Conference.
Unlike the PyTorch deep learning framework for research, experimentation, and exploration of CNNs, Caffe2 will support mobile industry re-use more. Of course, this does not mean that PyTorch does not support mobile applications, but Caffe2 has more examples of CNNs.
At the F8 conference, Facebook said that it is applying AI to the AR platform, and both voice recognition in Oculus and image recognition in Facebook applications have used AI.
The Amazon cloud computing service platform AWS added support for the Caffe2 framework in Machine Image Deep Learning (AMI), and Microsoft's Azure cloud computing platform also uses Caffe2 to support its Data Science Virtualizer (DSVM).
In addition, Nvidia's deep learning GPUs that support Facebook services can help people better understand Caffe 2. Qualcomm said that its Dragon NPE will make full use of Caffe 2, and Intel will also integrate the mathematical core library (MKL) into the Caffe 2 framework.
Currently, there are several deep learning frameworks that have been open sourced, including TensorFlow, Torch, Theano, Deeplearning4j, and MXNet, which are suitable for people's different purposes. In 2015, Google announced the open source TensorFlow, which caused a huge stir.
6.5Inch Coaxial Speaker,6.5 Inch Coaxial,6.5Inch Church Speaker,6.5 Inch Coaxial Speaker
Guangzhou BMY Electronic Limited company , https://www.bmy-speakers.com