US Patent Application 18021810. TRAINING DEVICE, TRAINING METHOD AND TRAINING PROGRAM simplified abstract
Contents
TRAINING DEVICE, TRAINING METHOD AND TRAINING PROGRAM
Organization Name
Nippon Telegraph and Telephone Corporation
Inventor(s)
Shinya Yamaguchi of Musashino-shi, Tokyo (JP)
Sekitoshi Kanai of Musashino-shi, Tokyo (JP)
TRAINING DEVICE, TRAINING METHOD AND TRAINING PROGRAM - A simplified explanation of the abstract
This abstract first appeared for US patent application 18021810 titled 'TRAINING DEVICE, TRAINING METHOD AND TRAINING PROGRAM
Simplified Explanation
The patent application describes a system that converts data into frequency components and uses adversarial learning to optimize a generator and two discriminators.
- The system converts first data into a first frequency component and second data generated by a generator into a second frequency component.
- A loss function is calculated to optimize the generator, a first discriminator, and a second discriminator.
- The first discriminator distinguishes between the first data and the second data, while the second discriminator distinguishes between the first frequency component and the second frequency component.
- Parameters of the generator, first discriminator, and second discriminator are updated to optimize the calculated loss function.
Original Abstract Submitted
The conversion unit () converts first data into a first frequency component, and converts second data generated by a generator that configures an adversarial learning model into a second frequency component. The calculation unit () calculates a loss function that simultaneously optimizes the generator, a first discriminator that configures the adversarial learning model and discriminates between the first data and the second data, and a second discriminator that configures the adversarial learning model and discriminates between the first frequency component and the second frequency component. The update unit () updates parameters of the generator, the first discriminator, and the second discriminator so that the loss function calculated by the calculation unit () is optimized.