In most Computer Vision applications, Deep Learning models achieve state-of-the-art performances. One drawback of is the large amount data needed to train models. Unfortunately, in many are difficult or expensive collect. Data augmentation can alleviate problem, generating new from a smaller initial dataset. Geometric and color space image methods increase accuracy but often not enough. More ad...

Protecting the intellectual property (IP) of deep neural network (DNN) models becomes essential and urgent with rapidly increasing cost DNN training. Fingerprinting is one promising IP protection method that queries suspicious specific fingerprint samples to infer verify by comparing predictions pre-defined labels. Based on utilizing unique features target models, various fingerprinting methods...

Applications of machine learning are subject to three major components that contribute the final performance metrics. Within category neural networks, and deep specifically, first two architecture for model being trained training approach used. This work focuses on third component, data used during training. The primary questions arise “what is in data” within matters?” looking into radio frequ...

BACKGROUND A large, multicenter, 10-year observational study is being conducted to compare the long-term safety and effectiveness of Natrelle silicone breast implants with saline implants or national norms. Study baseline data and surgical characteristics are reported here. METHODS Women seeking primary augmentation, revision-augmentation, primary reconstruction, or revision-reconstruction pa...