ALTERNATIVE SOFT LABEL GENERATION

Title:

ALTERNATIVE SOFT LABEL GENERATION

Document Type and Number:

WIPO Patent Application WO/2022/121684

Kind Code:

A1

Abstract:

An approach to identifying alternate soft labels for training a student model may be provided. A teaching model may generate a soft label for a labeled training data. The training data can be an acoustic file for speech or a spoken natural language. A pool of soft labels previously generated by teacher models can be searched at the label level to identify soft labels that are similar to the generated soft label. The similar soft labels can have similar length or sequence at the word phoneme, and/or state level. The identified similar soft labels can be used in conjunction with the generated soft label to train a student model.

More Like This:

WO/2023/222113	SPARSE PARAMETER UPDATING METHOD, TRAINING NODE, DEVICE, AND STORAGE MEDIUM
WO/2019/229117	METHOD FOR TRANSFERRING LEARNING FROM A FIRST CONVOLUTIONAL NEURAL NETWORK TO A SECOND CONVOLUTIONAL NEURAL NETWORK
WO/2020/092086	ARTIFICIAL NEURAL NETWORKS HAVING COMPETITIVE REWARD MODULATED SPIKE TIME DEPENDENT PLASTICITY AND METHODS OF TRAINING THE SAME

Inventors:

NAGANO TOHRU (JP)
FUKUDA TAKASHI (JP)
KURATA GAKUTO (JP)

Application Number:

PCT/CN2021/132681

Publication Date:

June 16, 2022

Filing Date:

November 24, 2021

Export Citation:

Click for automatic bibliography generation Help

Assignee:

IBM (US)
IBM CHINA CO LTD (CN)

International Classes:

G06N3/08; G06N3/04; G10L15/06

Foreign References:

US20190205748A1	2019-07-04
CN110852426A	2020-02-28
CN110459208A	2019-11-15
US20170032244A1	2017-02-02
US20200167677A1	2020-05-28

Attorney, Agent or Firm:

LIU, SHEN & ASSOCIATES (CN)

Download PDF:

View/Download PDF PDF Help

Previous Patent: IDENTIFYING DEPENDENCIES IN CONTROL SEQUENCE FOR EXECUTION ON HARDWARE ACCELERATOR

Next Patent: EDGE COMPUTING AUTONOMOUS VEHICLE INFRASTRUCTURE