Browse Prior Art Database

Method and System for Cognitive Augmentation of Voice Commands with Intention Inference in Smart Home Systems

IP.com Disclosure Number: IPCOM000249352D
Publication Date: 2017-Feb-20
Document File: 3 page(s) / 150K

Publishing Venue

The IP.com Prior Art Database

Abstract

A method and system is disclosed for cognitive augmentation of voice commands with intention inference in smart home systems.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

1

Method and System for Cognitive Augmentation of Voice Commands with Intention Inference in Smart Home Systems

Current smart home systems, relies greatly on voice commands that are usually short sentences with key words. The raw commands usually do not contain enough information for execution and looking for deeper level intention is always challenging. While there are various related applications in market that facilitate better services, improving home system commands is still essential.

Disclosed is a method and system for cognitive augmentation of voice commands with intention inference in smart home systems. The method cognitively enhances voice commands with intention inference, by using a command coordinator, and a broadcast controller. The command coordinator, recognizes and processes the commands separately into two different types such as raw commands and feedback commands. Raw commands given by users can be such as, but not limited to, turn on the light, play music, and turn on the TV. Feedback commands given by users can be such as, but not limited to, I mean nightstand lamp, too loud, and good job. The broadcast controller facilitates internet-of-things (IoTs) in the smart home, to receive broadcast signals and transmit instant status.

In accordance with the method and system, as illustrated in Figure 1, the raw commands and the IoT status are utilized for multi-modal reinforcement learning and generating augmented (modified) commands for transmitting back to the command coordinator and then to the broadcast controller to take real action. Alternatively, the feedback command and the IoT status are utilized for multi-modal reinforcement learning to enhance historical data pool for future reference. Subsequently, the historical data is utilized to create cognitive family member profiling and further enhancing data pool by individual routine and habit.

Figure 1

2

As illustrated in Figure 1, the IoT layer comprises IoT devices such as, but not limited to, phone, smar...