FTSwish with mean shifting added to increase performance. Original FTSwish is from this paper: https://arxiv.org/abs/1812.06247
FTSwishPlus = This is FTSwish, but with a mean shift added to drive the mean to zero on a base kaiming init'ed tensor.
This is to help ensure a more stable starting setup and the concept developed by FastAI/Jeremy Howard. By adding mean shift of - .1, performance improved so that it was able to win an initial comparison of Relu, General Relu, LiSHT for image classification.
More detailed article with testing results: https://medium.com/@lessw/comparison-of-activation-functions-for-deep-learning-initial-winner-ftswish-f13e2621847