US 12,170,779 B2
Training a data coding system comprising a feature extractor neural network
Francesco Cricri, Tampere (FI); Nam Le, Tampere (FI); Hamed Rezazadegan Tavakoli, Espoo (FI); Honglei Zhang, Tampere (FI); Miska Matias Hannuksela, Tampere (FI); and Emre Baris Aksu, Tampere (FI)
Assigned to Nokia Technologies Oy, Espoo (FI)
Appl. No. 17/917,153
Filed by Nokia Technologies Oy, Espoo (FI)
PCT Filed Mar. 30, 2021, PCT No. PCT/FI2021/050227
§ 371(c)(1), (2) Date Oct. 5, 2022,
PCT Pub. No. WO2021/205065, PCT Pub. Date Oct. 14, 2021.
Claims priority of application No. 20205380 (FI), filed on Apr. 9, 2020.
Prior Publication US 2023/0164336 A1, May 25, 2023
Int. Cl. H04N 19/146 (2014.01); G06N 3/0455 (2023.01); G06N 3/084 (2023.01); G06V 10/44 (2022.01); G06V 10/82 (2022.01); H04N 19/192 (2014.01); H04N 19/42 (2014.01)
CPC H04N 19/146 (2014.11) [G06N 3/0455 (2023.01); G06N 3/084 (2013.01); G06V 10/454 (2022.01); G06V 10/82 (2022.01); H04N 19/192 (2014.11); H04N 19/42 (2014.11)] 20 Claims
OG exemplary drawing
 
1. An apparatus, comprising:
at least one processor; and
at least one memory storing instructions that, when executed by the at least one processor, cause the apparatus to:
obtain a data coding pipeline comprising a feature extractor neural network configured to extract features from input data, an encoder neural network configured to encode extracted features from the feature extractor neural network, and a decoder neural network configured to reconstruct the input data based on output of the encoder neural network;
determine a plurality of losses for the coding pipeline, the plurality of losses corresponding to at least a plurality of tasks;
determine a weight update for at least a subset of the coding pipeline based on the plurality of losses, wherein the weight update is configured to reduce a number of iterations for fine-tuning the coding pipeline for at least one of the plurality of tasks; and
update at least the subset of the coding pipeline based on the weight update.