US 11,704,549 B2
Event-based classification of features in a reconfigurable and temporally coded convolutional spiking neural network
Peter Aj Van Der Made, Nedlands (AU); Anil S. Mankar, Laguna Hills, CA (US); Kristofor D. Carlson, Laguna Hills, CA (US); and Marco Cheng, Laguna Hills, CA (US)
Assigned to BrainChip, Inc., Laguna Hills, CA (US)
Filed by BrainChip, Inc., Laguna Hills, CA (US)
Filed on Jan. 14, 2022, as Appl. No. 17/576,103.
Application 17/576,103 is a continuation of application No. 16/938,254, filed on Jul. 24, 2020, granted, now 11,227,210.
Claims priority of provisional application 62/878,426, filed on Jul. 25, 2019.
Prior Publication US 2022/0138543 A1, May 5, 2022
This patent is subject to a terminal disclaimer.
Int. Cl. G06N 3/063 (2023.01); G11C 11/41 (2006.01); G11C 11/54 (2006.01); G06N 3/08 (2023.01); G06N 3/049 (2023.01); G06T 3/40 (2006.01)
CPC G06N 3/063 (2013.01) [G06N 3/049 (2013.01); G06N 3/08 (2013.01); G06T 3/4046 (2013.01); G11C 11/41 (2013.01); G11C 11/54 (2013.01)] 27 Claims
OG exemplary drawing
 
1. A system, comprising:
an inbound filter configured to receive spikes and select relevant spikes from the received spikes, wherein the relevant spikes are relevant to a neuron;
a memory configured to store kernels;
a packet collection module configured to collect the relevant spikes until a predetermined number of relevant spikes have been collected in a packet in the memory, and to organize the collected relevant spikes by spatial coordinates in the packet; and
a convolution neural processor configured to perform convolution in the memory.