US 12,079,695 B2
Scale-permuted machine learning architecture
Xianzhi Du, Mountain View, CA (US); Yin Cui, Mountain View, CA (US); Tsung-Yi Lin, Sunnyvale, CA (US); Quoc V. Le, Sunnyvale, CA (US); Pengchong Jin, Mountain View, CA (US); Mingxing Tan, Newark, CA (US); Golnaz Ghiasi, Mountain View, CA (US); and Xiaodan Song, Mountain View, CA (US)
Assigned to GOOGLE LLC, Mountain View, CA (US)
Filed by Google LLC, Mountain View, CA (US)
Filed on Oct. 1, 2020, as Appl. No. 17/061,355.
Prior Publication US 2022/0108204 A1, Apr. 7, 2022
Int. Cl. G06N 20/00 (2019.01); G06F 11/34 (2006.01); G06N 3/04 (2023.01)
CPC G06N 20/00 (2019.01) [G06F 11/3495 (2013.01); G06N 3/04 (2013.01)] 13 Claims
OG exemplary drawing
 
1. A computer-implemented method of generating scale-permuted models having improved accuracy or reduced computational requirements, the method comprising:
defining, by a computing system comprising one or more computing devices, a search space including a plurality of candidate permutations of a plurality of candidate feature blocks, each of the plurality of candidate feature blocks having a respective resolution;
performing, by the computing system, a plurality of search iterations by a search algorithm to select a scale-permuted model from the search space,
wherein the scale-permuted model comprises a sequence of blocks comprising:
a first feature block in the sequence having a first resolution,
a second feature block next in the sequence after the first feature block, the second feature block having a second resolution higher than the first resolution, and
a third feature block next in the sequence after the second feature block, the third feature block having a third resolution lower than the second resolution and different from the first resolution, and
wherein the scale-permuted model is based at least in part on a candidate permutation of the plurality of candidate permutations, the candidate permutation comprising a plurality of permuted feature blocks having a permuted ordering that differs from an initial ordering of the plurality of candidate feature blocks; and
providing, by the computing system, the scale-permuted model as an output.