WiMi Hologram Cloud Inc (WIMI)

[StockpickerCAPC]

WiMi Hologram Cloud Developed AI Voice Interactive Vehicle Robot.

Source
https://www.newstrail.com/wimi-hologram-cloud-developed-ai-voice-interactive-vehicle-robot/

November 23, 2022

The intelligent vehicle industry is growing rapidly; and autonomous driving has become a research hot spot. The in-vehicle intelligent computing platform is a necessary solution to realize autonomous driving. And that cannot be done without in-vehicle AI chips and voice interaction technology. Voice control is now a basic command in smart cars. In China, cars loaded with AI voice climbed yearly, from 49.82% in 2019 to 63.25% in 2020 and then to 86% in 2021. High-frequency use scenarios are navigation, music, air conditioning, and phone calls. It is believed that AI voice will become a must for cars in the future.

Recently, WiMi Hologram Cloud, Inc. (NASDAQ:WIMI) disclosed a patented technology, an interactive vehicle robot based on AI voice control. The central part of the robot’s face is provided with a display screen, and the upper part of the left and right sides is connected with a rotating arm. The rear side’s upper edge and the middle are provided with audio and an LED display screen, respectively. The robot is developed to target the AI voice-controlled interactive vehicle industry.

The voice interaction process consists of four parts:
- speech acquisition (SA),
The process of SA is to record voices, sampling, and coding of audio;
- automatic speech recognition (ASR),
ASR is the conversion of speech information to machine-recognizable text information;
- natural language processing (NLP),
NLP is the corresponding operations based on the text characters or commands converted by speech recognition,
- text-to-speech (TTS).
TTS completes the conversion of text information to sound data.


Speech interaction system architecture:
The user interface connects the user and the system, and the characteristics of both the user and the system determine the design of the user interface. The speech interaction system contains three main parts, carried out sequentially to complete a round of human-computer interaction.
- SA and ASR:
The system acquires the user’s speech through a microphone array and converts it into text. The in-vehicle microphone receives the sound signal, converts it into an audio signal, and after decoding, noise canceling, audio enhancement, feature extraction, phoneme selection, and decoding again, it finally becomes High-levelspeech recognition will localize the sound source.
- NLP:
The systeminterprets the semantic meaning and generates a reply. After the voice signal is converted into text and to be recognized, the system picks up the corresponding functional vocabulary and generates commands. After filtering against a thesaurus and deep learning, the system picks up the instructions.
TTS:
The system converts the reply information into speech and gives feedback to the user. The system converts the recognized summary into command planning and sends it to the corresponding vehicle control unit to execute the action. Finally, the speech synthesis module converts the text of the completed action execution into speech broadcast feedback to the driver.

There are three significant technical difficulties in voice interaction: first, accurate voice pickup and noise cancellation; second, training data for speech deep learning model; and third, training of recognition feature extraction model. WiMi’s R&D team is vigorously developing AI voice interaction technology to improve the voice system of intelligent cars to achieve more efficient AI interaction functions.

WiMi is also developing and designing voice chips for AI voice interaction vehicle robots. The ultra-low-power far-field voice interaction chip designed for in-vehicle voice interaction scenarios has far-field array signal real-time processing, advanced ultra-low false alarm voice wake-up, offline voice recognition, and other functions. The launch of WiMi’s AI voice robot will also promote the growth of WiMi’s semiconductor business.

With the rise of AI, autonomous driving, voice recognition, and gesture interaction have gradually matured. AI now has permeated vehicles, education, and medical care. As an essential means of human-computer exchange, AI voice will also be widely used in various fields. In addition to everyday car scenes, AI voice can be applied in education and training, medical scenes, and government services.