Select one of plurality of neural networks
D Pye, C Waple - US Patent 11,113,596, 2021 - Google Patents
Data is input to one of a plurality of neural networks. Each of the plurality of neural networks
is to be of a different size. A propagation time is determined for the inputted data. The …
is to be of a different size. A propagation time is determined for the inputted data. The …
Method for optimizing an artificial neural network (ANN)
S Yao - US Patent 10,621,486, 2020 - Google Patents
US10621486B2 - Method for optimizing an artificial neural network (ANN) - Google Patents
US10621486B2 - Method for optimizing an artificial neural network (ANN) - Google Patents …
US10621486B2 - Method for optimizing an artificial neural network (ANN) - Google Patents …
Bit width selection for fixed point neural networks
D Lin, VSR Annapureddy, DJ Julian… - US Patent …, 2019 - Google Patents
(57) ABSTRACT A method for selecting bit widths for a fixed point machine learning model
includes evaluating a sensitivity of model accuracy to bit widths at each computational stage …
includes evaluating a sensitivity of model accuracy to bit widths at each computational stage …
Systems and methods for automatically generating code for deep learning systems
G Venkataramani, RP Kokku, J Shankar… - US Patent …, 2018 - Google Patents
Abstract Systems and methods may automatically generate code for deep learning
networks. The systems methods may provide a code generation framework for generating …
networks. The systems methods may provide a code generation framework for generating …
Coordination and increased utilization of graphics processors during inference
AR Appu, A Koker, JC Weast, MB MacPherson… - US Patent …, 2019 - Google Patents
A mechanism is described for facilitating inference coordination and processing utilization
for machine learning at autonomous machines. A method of embodiments, as described …
for machine learning at autonomous machines. A method of embodiments, as described …
Neural network architecture for analyzing video data
MW Ferro, M Woodward - US Patent App. 15/382,438, 2017 - Google Patents
Embodiments are provided for analyzing and characterizing video data. According to certain
aspects, an analysis machine may analyze video data and optional audio data …
aspects, an analysis machine may analyze video data and optional audio data …
Object recognition from videos using recurrent neural networks
A Angelova, I Bogun - US Patent 10,013,640, 2018 - Google Patents
Methods, systems, and apparatus, including computer programs encoded on computer
storage media, for identifying an object from a video. One of the methods includes obtaining …
storage media, for identifying an object from a video. One of the methods includes obtaining …
Neural network processor using compression and decompression of activation data to reduce memory bandwidth utilization
JL Corkery, BE Lundell, LM Wall, CB McBRIDE… - US Patent …, 2022 - Google Patents
(57) ABSTRACT A deep neural network (“DNN”) module compresses and decompresses
neuron-generated activation data to reduce the utilization of memory bus bandwidth. The …
neuron-generated activation data to reduce the utilization of memory bus bandwidth. The …
Method and apparatus for generating fixed-point type neural network
H Yim, DY Kim, B Kim, N Sung, JH Lim… - US Patent 11,373,087, 2022 - Google Patents
A method of generating a fixed-point type neural network by quantizing a floating-point type
neural network, includes obtaining, by a device, a plurality of post-activation values by …
neural network, includes obtaining, by a device, a plurality of post-activation values by …
Multi-modal construction of deep learning networks
Methods, systems, and computer program products for multi-modal construction of deep
learning networks are provided herein. A computer-implemented method includes …
learning networks are provided herein. A computer-implemented method includes …