US 11,698,947 B2
Empathic artificial intelligence systems
Alan Cowen, New York, NY (US); Dacher Keltner, Berkeley, CA (US); and Bill Schoenfeld, Tokyo (JP)
Assigned to Hume AI Inc., New York, NY (US)
Filed by Hume AI Inc., New York, NY (US)
Filed on Oct. 13, 2022, as Appl. No. 17/965,375.
Application 17/965,375 is a continuation of application No. 17/742,246, filed on May 11, 2022, granted, now 11,551,031.
Claims priority of provisional application 63/209,870, filed on Jun. 11, 2021.
Prior Publication US 2023/0096485 A1, Mar. 30, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. G06F 18/40 (2023.01); A61B 5/16 (2006.01); G16H 50/20 (2018.01); G06F 18/214 (2023.01); G06N 20/00 (2019.01)
CPC G06F 18/41 (2023.01) [A61B 5/165 (2013.01); G06F 18/2148 (2023.01); G16H 50/20 (2018.01); G06N 20/00 (2019.01)] 30 Claims
OG exemplary drawing
 
1. A method for obtaining training data for predicting expressions from received media data comprising:
displaying a predefined media content and a plurality of predefined expression tags;
receiving, from a user, a selection of one or more expression tags from the plurality of predefined expression tags;
receiving a recording of the user imitating the predefined media content;
storing the recording in association with the selected one or more expression tags as training data; and
training a machine learning model, the machine-learning model configured to receive input media data and predict an expression based on the input media data.
 
19. A method for predicting one or more expressions in a media data, comprising:
receiving the media data;
inputting the received media data into a trained machine-learning model, the machine-learning model trained by a process comprising:
displaying a predefined media content and a plurality of predefined expression tags;
receiving, from a user, a selection of one or more expression tags from the plurality of predefined expression tags;
receiving a recording of the user imitating the predefined media content; and
storing, as training data, the recording in association with the selected one or more expression tags; and
predicting the one or more expressions in the received media data based on an output of the trained machine-learning model.
 
23. A system for predicting one or more expressions in a media data, comprising:
one or more processors; and
a memory communicatively coupled to the one or more processors and configured to store instructions that, when executed by the one or more processors, cause the system to:
receive the media data;
input the received media data into a trained machine-learning model, the machine-learning model trained by a process comprising:
displaying a predefined media content and a plurality of predefined expression tags;
receiving, from a user, a selection of one or more expression tags from the plurality of predefined expression tags;
receiving a recording of the user imitating the predefined media content; and
storing, as training data, the recording in association with the selected one or more expression tags; and
predict the one or more expressions in the received media data based on an output of the trained machine-learning model.