Neural Networks have been shown to be sensitive to common perturbations such as blur, Gaussian noise, rotations, etc. They are also vulnerable to some artificial malicious corruptions called adversarial examples. The adversarial examples study has recently become very popular and it sometimes even reduces the term "adversarial robustness" to the term "robustness". Yet, we do not know to what extent the adversarial robustness is related to the global robustness. Similarly, we do not know if a robustness to various common perturbations such as translations or contrast losses for instance, could help with adversarial corruptions. We intend to study the links between the robustnesses of neural networks to both perturbations. With our experiments, we provide one of the first benchmark designed to estimate the robustness of neural networks to common perturbations. We show that increasing the robustness to carefully selected common perturbations, can make neural networks more robust to unseen common perturbations. We also prove that adversarial robustness and robustness to common perturbations are independent. Our results make us believe that neural network robustness should be addressed in a broader sense.

Researchers at MIT have developed a new hardware chip that can make neural networks in smart devices more power efficient. Usually, neural networks require much energy as they are computing a large of data at once, so while there are neural features on devices such as smartphones, they are integrated on a smaller scale. The data is usually is sent to the web instead to process all the information and sent back to the device. However, these chips compute information three to seven times faster than currently dedicated chips. Not only that, they reportedly consume 94 to 95 percent less power, making them an idea inclusion on smaller handheld devices in the future.