# US 2023/0412655 A1 · Systems and methods for communication monitoring

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2023/0412655 A1
- Document ID: US20230412655A1
- Status: Published application
- Kind: A1
- Application number: 17/828,907
- Filed: 2022-05-31
- Published: 2023-12-21
- Earliest claimed: 2022-05-31
- Assignees: Avaya Management L.P.
- Applicants: Avaya Management L.P.
- Inventors: Jinson Abraham
- Web page: https://patentgrep.com/patent/US20230412655A1

## CPC

- H04L 65/403: Arrangements for multi-party communication, e.g. for conferences (data switching systems for conference H04L12/18; arrangements for connecting several subscribers to a common circuit, i.e. affording conference facilities H04M3/56; television conferencing systems H04N7/15)
- G06F 40/10: Text processing (natural language analysis G06F40/20; semantic analysis G06F40/30; processing or translation of natural language G06F40/40)
- G06F 40/30: Semantic analysis
- H04L 65/1089: by adding media; by removing media
- H04L 65/1093: by adding participants; by removing participants
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

In a traditional classroom setting a teacher is present to monitor the student and their behavior. With virtual classrooms becoming more common, and parents often required to be physically present at their jobs or other engagements, students are often home alone. A teacher conducting a virtual class may not notice when a student closes their eyes, leaves, or performs other undesired actions. By providing the students with a full two-way communication with their teacher, the student is presented with a lesson and classroom activities. By providing a parent or other authorized person with a feed signal that excludes children that are not their own, the parent may remotely monitor the student, such as to ensure the student is attentive to their teacher or otherwise not engaged in any undesired activities.

## Claims

### Claim 1

A system for monitoring a first device comprising:
a network interface to a network;
a processor comprising instructions maintained in a non-transitory storage; and
wherein the processor performs:
hosting a first communication, via the network, between a second device and a plurality of endpoints wherein the first device is one of the plurality of endpoints and wherein each of the plurality of endpoints provides a feed signal as a portion of content of the first communication; and
transmitting, via the network, a second communication to a third device, the second communication occurring concurrently with the first communication and comprising content from the content of the feed signal that is provided by the first device alone and wherein the second communication omits content from each of the plurality of endpoints other than the first device.

### Claim 2 (depends on 1)

The system of claim 1, wherein the processor further performs:
monitoring the feed signal from the first device;
determining the feed signal comprises undesired content;
generating an alert comprising indicia of the undesired content; and
sending the alert to the third device.

### Claim 3 (depends on 2)

The system of claim 2, wherein the processor determines the feed signal comprises undesired content further comprising:
providing the feed signal to a neural network trained to detect undesired content; and
receiving a decision from the neural network that undesired content is present in the feed signal.

### Claim 4 (depends on 3)

The system of claim 3, wherein:
the neural network is trained to receive the feed signal and return a decision whether undesirable content is present in the feed signal, comprising:
collecting a set of undesired content from a database;
applying one or more transformations to each undesired content including altering a first spoken word to a second spoken word, a first typed word to a second typed word, inserting a word into a string of words, replacing a first body position with a second body position, replacing a first eye position with a second eye position, and replacing a first degree of physical body motion with a second degree of physical body motion to create a modified set of undesired content;
creating a first training set comprising the collected set of undesired content, the modified set of undesired content, and a set of desired content;
training the neural network in a first stage using the first training set;
creating a second training set for a second stage of training comprising the first training set and desired content that are incorrectly detected as undesired content after the first stage of training; and
training the neural network in the second stage using the second training set.

### Claim 5 (depends on 3)

The system of claim 3, wherein:
the neural network is trained to receive the feed signal and return a decision whether undesirable content is present in the feed signal, comprising:
collecting a set of desired content from a database;
applying one or more transformations to each desired content including altering a first spoken word to a second spoken word, a first typed word to a second typed word, inserting a word into a string of words, replacing a first body position with a second body position, replacing a first eye position with a second eye position, altering an article of clothing, and replacing a first degree of physical body motion with a second degree of physical body motion to create a modified set of desired content;
creating a first training set comprising the collected set of desired content, the modified set of desired content, and a set of undesired content;
training the neural network in a first stage using the first training set;
creating a second training set for a second stage of training comprising the first training set and undesired content that are incorrectly detected as desired content after the first stage of training; and
training the neural network in the second stage using the second training set.

### Claim 6 (depends on 3)

The system of claim 3, wherein:
the neural network is trained to receive the feed signal and return a decision whether undesirable content is present in the feed signal, comprising:
accessing a set of historical feed signals previously received from a prior set of endpoints engaged in a prior first communication; and
receiving indicia from a historic third device indicating undesired content associated with a portion of one or more historical feed signals.

### Claim 7 (depends on 1)

The system of claim 1, wherein the processor further performs:
receiving a request to create the first communication;
in response to receiving the request, the processor generates a first link to an address to join the first communication and a second link to an address to join the second communication; and
sending a first invitation comprising the first link to a first address associated with a first party and a second invitation comprising the second link to a second address associated with a second party.

### Claim 8 (depends on 1)

The system of claim 1, wherein the processor further performs:
receiving a request from the first device to disable video from the first device; and
in response to receiving the request, omitting a video portion of the feed signal from the first communication and maintaining the video portion of the feed signal in the second communication.

### Claim 9 (depends on 1)

The system of claim 1, wherein the processor further performs providing indicia of attendance of the second device when the second device is receiving the second communication.

### Claim 10 (depends on 1)

The system of claim 1, wherein the processor further performs:
hosting a third communication, via the network, between a third device and the plurality of endpoints wherein the third device is one of the plurality of endpoints;
determining the second device is authorized to receive the first communication and the third communication; and
transmitting, via the network, a fourth communication to the third device, the fourth communication occurring concurrently with the second communication and comprising content from the content of the feed signal that is provided by the third device.

### Claim 11

A method for monitoring communications comprising a first device and conducted over a network comprising:
hosting a first communication, via the network, between a second device and a plurality of endpoints wherein the first device is one of the plurality of endpoints and wherein each of the plurality of endpoints provides a feed signal as a portion of content of the first communication; and
transmitting, via the network, a second communication to a third device, the second communication occurring concurrently with the first communication and comprising content from the content of the feed signal that is provided by the first device alone and wherein the second communication omits content from each of the plurality of endpoints other than the first device.

### Claim 12 (depends on 11)

The method of claim 11, further comprising:
monitoring the feed signal from the first device;
determining the feed signal comprises undesired content;
generating an alert comprising indicia of the undesired content; and
sending the alert to the third device.

### Claim 13 (depends on 12)

The method of claim 12, wherein determining the feed signal comprises undesired content further comprising:
providing the feed signal to a neural network trained to detect undesired content; and
receiving a decision from the neural network that undesired content is present in the feed signal.

### Claim 14 (depends on 13)

The method of claim 13, further comprising:
the neural network is trained to receive the feed signal and return a decision whether undesirable content is present in the feed signal, the training comprising:
collecting a set of undesired content from a database;
applying one or more transformations to each undesired content including altering a first spoken word to a second spoken word, a first typed word to a second typed word, inserting a word into a string of words, replacing a first body position with a second body position, replacing a first eye position with a second eye position, and replacing a first degree of physical body motion with a second degree of physical body motion to create a modified set of undesired content;
creating a first training set comprising the collected set of undesired content, the modified set of undesired content, and a set of desired content;
training the neural network in a first stage using the first training set;
creating a second training set for a second stage of training comprising the first training set and desired content that are incorrectly detected as undesired content after the first stage of training; and
training the neural network in the second stage using the second training set.

### Claim 15 (depends on 13)

The method of claim 13, further comprising:
the neural network is trained to receive the feed signal and return a decision whether undesirable content is present in the feed signal, the training comprising:
collecting a set of desired content from a database;
applying one or more transformations to each desired content including altering a first spoken word to a second spoken word, a first typed word to a second typed word, inserting a word into a string of words, replacing a first body position with a second body position, replacing a first eye position with a second eye position, and replacing a first degree of physical body motion with a second degree of physical body motion to create a modified set of desired content;
creating a first training set comprising the collected set of desired content, the modified set of desired content, and a set of undesired content;
training the neural network in a first stage using the first training set;
creating a second training set for a second stage of training comprising the first training set and undesired content that are incorrectly detected as desired content after the first stage of training; and
training the neural network in the second stage using the second training set.

### Claim 16 (depends on 13)

The method of claim 13, wherein:
the neural network is trained to receive the feed signal and return a decision whether undesirable content is present in the feed signal, the training comprising:
accessing a set of historical feed signals previously received from a prior set of endpoints engaged in a prior first communication; and
receiving indicia from a historic third device indicating undesired content associated with a portion of one or more historical feed signals.

### Claim 17 (depends on 13)

The method of claim 13, further comprising:
receiving a request to create the first communication;
in response to receiving the request, generating a first link to an address to join the first communication and a second link to an address to join the second communication; and
sending a first invitation comprising the first link to a first address associated with a first party and a second invitation comprising the second link to a second address associated with a second party.

### Claim 18 (depends on 11)

The method of claim 11, further comprising:
receiving a request from the first device to disable video from the first device; and
in response to receiving the request, omitting a video portion of the feed signal from the first communication and maintaining the video portion of the feed signal in the second communication.

### Claim 19 (depends on 11)

The method of claim 11, further comprising providing indicia of attendance of the second device when the second device is receiving the second communication.

### Claim 20

A system for monitoring a first device, comprising:
means to hosting a first communication between a second device and a plurality of endpoints wherein the first device is one of the plurality of endpoints and wherein each of the plurality of endpoints provides a feed signal as a portion of content of the first communication; and
means to transmit a second communication to a third device, the second communication occurring concurrently with the first communication and comprising content from the content of the feed signal that is provided by the first device alone and wherein the second communication omits content from each of the plurality of endpoints other than the first device.
