We just released the SignalPop AI Designer, a comprehensive AI development environment for Windows C# developers using the MyCaffe AI Platform! With the AI Designer, you can easily develop your own AI models right on your development computer using all of the development tools that you are familiar with!
All that you need to start deep learning is a Windows 7 or 10 based PC with a NVIDIA GPU, Microsoft SQL Express (which is free), and the SignalPop AI Designer and you are ready to go – see the SignalPop AI Designer Product page for more details on the minimum requirements needed to start building AI. Once your models are built they are ready to be used with the open-source MyCaffe AI platform which is easily integrated into your Visual Studio projects via the MyCaffe package on NuGet.
Now you can manage your datasets from CIFAR-10 to MNIST to your own custom datasets. The T-SNE algorithm integrated into the AI Designer then allows you to visualize how learnable each data set is. And with the debugging features of the AI Designer, you can debug models as they train in real-time which can help diagnose why your model is not performing as expected. In addition, several of the included model evaluators allow you to visualize trained models using Deconvolution and Unpooling or using the DeepDraw and DeepDream algorithms which give you a view of what the network actually sees.
For more information on this product, log in and take a look at the Getting Started document which describes most of the product features in detail.
In addition, check out the open-source MyCaffe project on GitHub used to build the underlying AI platform that the SignalPop AI Designer uses.
Also, see us on Nuget.org by searching for the MyCaffe package.
Time to start (deep) learning!
Using the SignalPop AI Designer (soon to be released) we imported the GoogleNet model, (originally trained by Berkeley Artificial Intelligence Research) and boot-strapped our training of the CIFAR-10 dataset up to 88% in about 30 minutes with the SignalPop AI Designer. Next, we generated the images from the 10 CIFAR-10 classes using the SignalPop AI Designer’s Dream Evaluator and the MyCaffe implementation of the Deep Draw algorithm (inspired by the Google deep dream algorithm). We then synchronized the deep draw images with great electronic dance music by CLOV15.
For more cool examples created with the SignalPop AI Designer, see our Examples section.
Sign up for a free account to get access to white papers, the free SignalPop AI Designer download and more!