Personalized lecture generator

Patent metadata

Field	Value
Publication number	US-12567340-B2
Application number	US-202218068447-A
Country	US
Kind code	B2
Filing date	Dec 19, 2022
Priority date	Dec 20, 2021
Publication date	Mar 3, 2026
Grant date	Mar 3, 2026

Abstract

In a first aspect, the present disclosure relates to a computer-implemented method for generating a personalized lecture for a user. The method comprises obtaining at least one time series of electrodermal activity data of a user from a sensor and identifying a mental state of the user based on the at least one time series of electrodermal activity data. The method further comprises obtaining teaching content and generating a personalized lecture based on the mental state and the teaching content.

First claim

The invention claimed is:

1 . A computer-implemented method for generating a personalized lecture video sequence for a user, comprising:
obtaining at least one time series of electrodermal activity data of the user from a sensor; identifying a mental state of the user based on the at least one time series of the electrodermal activity data; obtaining a teaching content; and generating the personalized lecture video sequence based on the mental state and the teaching content, wherein the personalized lecture video sequence includes a person having generated mouth movements based on a related audio sequence, and further wherein the generated mouth movements are generated by a neural network trained on at least one prior series of real mouth movements by the person and a corresponding predetermined audio sequence.

2 . The computer-implemented method of claim 1 , wherein obtaining the teaching content comprises capturing at least one time series of audio data by a microphone.

3 . The computer-implemented method of claim 2 , wherein the at least one time series of the audio data is interpreted into alphanumerical form to generate a time series of alphanumerical data.

4 . The computer-implemented method of claim 1 , wherein the at least one time series of the electrodermal activity data is captured by the sensor configured to measure electric conductivity of skin.

5 . The computer-implemented method of claim 1 , wherein the sensor is comprised within a writing instrument.

6 . The computer-implemented method of claim 1 , wherein the at least one time series of the electrodermal activity data comprises at least one time series of a skin conductance level and/or a skin conductance response.

7 . The computer-implemented method of claim 1 , wherein obtaining the teaching content comprises a teaching content identification algorithm, wherein the teaching content identification algorithm comprises:
determining at least one text module of at least one time series of alphanumerical data, wherein the at least one time series of the alphanumerical data comprise N-grams, bi-grams, Noun phrases, themes and/or facets; and obtaining the teaching content conveyed by comparing the at least one text module to predetermined text modules.

8 . The computer-implemented method of claim 1 , wherein the neural network comprises a generative adversarial network framework.

9 . The computer-implemented method of claim 1 , wherein identifying mental state of the user based on the at least one time series of the electrodermal activity data comprises applying a mental state recognition algorithm.

10 . The computer-implemented method of claim 9 , wherein the mental state recognition algorithm is implemented by a recurrent neural network that processes the at least one time series of electrodermal activity data, and wherein the mental state recognition algorithm processes the at least one series of electrodermal activity data by a applying a trough-to-peak technique and/or a continuous decomposition analysis.

11 . The computer-implemented method of claim 10 , wherein the at least one time series of the electrodermal activity data comprises at least one time series of a skin conductance level, and wherein the skin conductance level is a floating average of the skin conductance over a time period between about 2 minutes to about 20 minutes.

12 . The computer-implemented method of claim 11 , wherein the identified mental state is ranked into at least one of a low tier or a high tier.

13 . The computer-implemented method of claim 1 , wherein an alertness signal is generated depending on the identified mental state.

14 . The computer-implemented method of claim 1 , wherein generating the personalized lecture comprises applying a personalized lecture content creation algorithm configured to generate a personalized lecture content based on the teaching content, wherein the personalized lecture content creation algorithm comprises querying a lecture content database based on the teaching content, in particular querying the lecture content database based on the teaching content to obtain data related to the teaching content, and wherein the personalized content creation algorithm comprises processing the teaching content or the data related to the teaching content by an autoregressive language model, wherein the autoregressive language model uses deep learning comprising a generative pre-trained transformer.

15 . The computer-implemented method of claim 14 , wherein the personalized lecture comprises a personalized lecture audio sequence based on the personalized lecture content, in particular a personalized lecture audio sequence of a lecturer based on the personalized lecture content.

16 . A computer system for generating a personalized lecture video sequence for a user, comprising:
one or more processors; and at least one non-transitory computer readable medium storing instructions which, when executed by the one or more processors, cause the one or more processors to perform operations comprising:
obtaining at least one time series of electrodermal activity data of the user from a sensor;
identifying a mental state of the user based on the at least one time series of the electrodermal activity data;
obtaining a teaching content; and
generating the personalized lecture video sequence based on the mental state and the teaching content, wherein the personalized lecture video sequence includes a person having generated mouth movements based on a related audio sequence, and further wherein the generated mouth movements are generated by a neural network trained on at least one prior series of real mouth movements by the person and a corresponding predetermined audio sequence.

17 . The computer system of claim 16 , comprising:
a writing instrument comprising:
the sensor configured to measure electric conductivity of skin; and
a first interface;
a processing unit remote from the writing instrument, comprising:
a second interface, wherein the first and second interface are configured to exchange data.

18 . A writing instrument comprising:
a sensor configured to measure electric conductivity of skin; a first interface configured to communicate and exchange data with a second interface of a processing unit, wherein the processing unit is configured to perform operations comprising:
obtaining at least one time series of electrodermal activity data of a user of the writing instrument from the sensor;
identifying a mental state of the user based on the at least one time series of the electrodermal activity data;
obtaining a teaching content; and
generating a personalized lecture video sequence based on the mental state and the teaching content, wherein the personalized lecture video sequence includes a person having generated mouth movements based on a related audio sequence, and further wherein the generated mouth movements are generated by a neural network trained on at least one prior series of real mouth movements by the person and a corresponding predetermined audio sequence.

19 . The writing instrument of claim 18 , wherein the writing instrument comprises a signaler, and wherein the signaler comprises a vibrating alert configured to activate upon receiving an alertness signal.

20 . A non-transitory computer-readable medium for generating a personalized lecture video sequence for a user, the non-transitory computer-readable medium storing instructions which, when executed by one or more processors, cause

Assignees

• Bic Violex Single Member Sa

Inventors

• Kokozidis Michail

Classifications

• G09B5/06 Primary

  with both visual and audible presentation of the material to be studied · CPC title

• G08B6/00

  Tactile signalling systems, e.g. tactile personal calling systems · CPC title

• A61B5/165

  Evaluating the state of mind, e.g. depression, anxiety · CPC title

• A61B5/0531

  Measuring skin impedance · CPC title

• G06N3/0475

  Generative networks · CPC title