18884321. Neural Network with Approximated Activation Function (Huawei Technologies Co., Ltd.)
Neural Network with Approximated Activation Function
Organization Name
Inventor(s)
Timofey Mikhailovich Solovyev of Munich (DE)
Alexander Alexandrovich Karabutov of Munich (DE)
Maxim Borisovitch Sychev of Moscow (RU)
Mikhail Vyacheslavovich Sosulnikov of Moscow (RU)
Sergey Yurievich Ikonin of Moscow (RU)
Elena Alexandrovna Alshina of Munich (DE)
Neural Network with Approximated Activation Function
This abstract first appeared for US patent application 18884321 titled 'Neural Network with Approximated Activation Function
Original Abstract Submitted
A neural network including at least one neural network layer and an activation function connected to an output of the at least one neural network layer. The activation function is implemented as an approximation function of a mathematically defined real valued non-linear activation function, wherein the approximation function allows for integer-only processing of fixed-point representations of input values of the approximation function.
- Huawei Technologies Co., Ltd.
- Timofey Mikhailovich Solovyev of Munich (DE)
- Esin Koyuncu of Munich (DE)
- Alexander Alexandrovich Karabutov of Munich (DE)
- Maxim Borisovitch Sychev of Moscow (RU)
- Mikhail Vyacheslavovich Sosulnikov of Moscow (RU)
- Sergey Yurievich Ikonin of Moscow (RU)
- Elena Alexandrovna Alshina of Munich (DE)
- H04N19/189
- H04N19/13
- H04N19/42
- CPC H04N19/189