Application of IoT-Enabled CNN for Natural Language Processing
Ashish Kumar, Rishab Mamgai, Rachna Jain
Abstract
Internet of Things (IoT) based systems are used to define communication systems based on machine-to-machine interaction. IoT when integrated with convolution neural network (CNN) can provide a system that can communicate with surroundings using human speech. Natural language processing (NLP) can interact with IoT-based deep learning systems to provide development in the automation field. IoT can connect a network of specific devices and exploit deep learning for feature extraction, namely sensor features, radio frequency features, and speech features. IoT with NLP can develop speech-based recognition systems for home automation systems. Smart home applications can be integrated with voice-command-based IoT devices to communicate specific commands to the devices. In addition, NLP-based IoT devices can help disabled people to perform their daily activities. These devices can monitor their health and provide voice-based security alerts. Also, NLP-enabled IoT devices can be helpful for automating environmental data collections which include geographical activities. However, NLP-based IoT implementation has certain limitations, namely language understanding, change in accent, and change in voice. These challenges restrict the efficient and quick utilization of NLP-based IoT devices. The deep learning technology with a big vocabulary database has provided numerous opportunities to 150train the voice and command recognition system in IoT. IoT-enabled CNN devices for voice recognition act as a boon to society.