What You Don’t Know About AI Voice Assistants

What You Don’t Know About AI Voice Assistants

What is a Smart Voice Assistant?

In this rapidly changing era, we may have already experienced overwhelming promotions about “voice assistants”. As you read this article, you might be holding a smartphone with a “smart voice assistant” and letting it gather dust in the corner of your phone.

A voice assistant is an intelligent mobile application that helps users solve problems through intelligent dialogue and instant Q&A interactions. Its main purpose is to assist users with daily life issues. Apple’s Siri pioneered the smart voice assistant, and of course, Chinese voice assistants have flourished like bamboo shoots after a rain, with their intelligent search capabilities in Chinese even surpassing Siri.

iFLYTEK has ranked first in various domestic and international evaluations for Chinese voice technology. Their requirements for mobile voice assistants are: “In addition to being able to listen and speak, they must also understand and think.” For instance, Google’s Assistant in mobile chat scenarios can intelligently analyze dialogues between two people and quickly provide responses and suggestions.

At the launch of Siri, although it supported few languages, had slow responses, and poor semantic understanding, people were still eager to try its features out of curiosity.

As time passed, Siri added more features and improved application performance, connecting with phone, maps, reminders, music, and other functions to handle simple tasks for users at appropriate times.

Application Location: On Apple Watch.

Commonly Used Features: Create reminders, set timers, etc.

In 2008, Google first launched a voice assistant, and in 2012 established a knowledge graph that allows searching for everything Google knows about people, events, and things, making search intelligent. Now with the launch of Google Assistant, it has contextual awareness capabilities, which not only understand the context of questions but also analyze the true intent behind the questions.

In 2016, Google released a hardware device called Google Home and an IM software called Allo equipped with Assistant.

Google Allo, an intelligent instant messaging application, focuses on artificial intelligence and Google Assistant, capable of automatically replying to messages.

Microsoft Xiaoice & Cortana

Microsoft Xiaoice is the world’s first personal intelligent assistant released by Microsoft. It “can understand users’ preferences and habits” and “help users with scheduling, answering questions, etc.” It records users’ behaviors and usage habits, analyzing unstructured data using cloud computing and search engines, to read and “learn” from text files, emails, images, videos, etc., on the phone, thus achieving human-computer interaction.

Baidu Dumi

Baidu Dumi has a wide coverage and understands Chinese context better, making interactions more convenient and better suited for Chinese users, but it still needs to strengthen globalization.

Lingxi Voice Assistant

The Lingxi Voice Assistant has the highest recognition rate for Mandarin and is the leading Chinese voice assistant in the domestic market; jointly launched by iFLYTEK, a leader in Chinese voice technology, and China Mobile, it not only adopts the world’s most advanced voice recognition technology but also offers rich localized services, with accurate recognition and rapid wake-up, especially optimized for recognizing Chinese accents.

Bixby

Bixby, an artificial intelligence assistant, is a smart voice control feature used in Samsung Galaxy S8 and S8+. It is a research result combining Samsung’s artificial intelligence, deep learning, and UI design. Bixby has the ability to understand context, allowing users to complete operations continuously. Applications equipped with Bixby will have contextual understanding capabilities, enabling them to more accurately understand what users want to do and can understand and execute voice commands even when the information is incomplete or the sentence order is incorrect. Bixby can understand incomplete instructions or those that do not fully match presets, making interaction more natural and easier.

Essentially, they are not much different, mainly depending on the usage scenario:

If you want to use voice to perform operations on your phone, Google’s Assistant and Microsoft’s Xiaoice are great choices for tasks like hailing a ride, posting on social media, or setting alarms. Google’s Assistant can even help you toggle hotspot settings. If you just want to have fun chatting with an interesting app, both Microsoft’s Xiaoice and Apple’s Siri are good options, especially Microsoft’s Xiaoice, which can even imitate Zhao Benshan and maintain semantic links between conversations rather than just recognizing the current sentence.

Why are hardware manufacturers rushing to deploy smart voice technology? Clearly, simple explanations like “herd mentality” or “gimmick theory” do not hold water.

In the past, voice communication has become a focal point for people’s attention, but why has this technology only now become a major issue in this decade? Despite the fact that computers have been able to accept voice commands since the 1990s, this connection has never fully realized its true potential. Now, thanks to the development of cloud data and artificial intelligence, we finally have the capability to achieve a truly conversational interface.

At a conference in 2017, Google claimed it was using deep learning algorithms to reduce its word error rate from 8.9% to 4.5% within a year. This progress in voice processing technology brings great hope for voice user interfaces. This is good news for any industry that can benefit from voice interaction.

Let’s take a look at the industries where voice technology will fundamentally change how consumers use products and services.

Automobiles

Many drivers are already using voice communication on their phones while driving. If the automotive industry acts quickly, it could become the primary interface for users, especially for those who spend hours stuck in traffic every day. In the future, driving will be a completely different experience, and your car might soon become a layer of interface in your life.

Currently, the things drivers can do while driving are very limited. Due to safety regulations, the automotive industry can only offer limited tactile interactions. A hands-free voice interface will provide car manufacturers with many new options. This technology can also keep drivers’ eyes on the road, thus enhancing safety.

High-Tech Wearables

Due to voice technology, screenless wearable devices will continue to succeed. So far, devices like Fitbit have become sensors that require additional mobile interfaces. With voice capabilities, these wearable devices can begin to operate independently of operating systems.

Customer Service

Voice technology has greatly impacted customer service, and conversational interfaces will bring even more changes in the coming years. One example is that call centers have been using voice recognition technology to enhance their conversational abilities. Any role that relies on communication will improve through voice interfaces. While human-to-human contact is always ideal, in many cases, voice interfaces can strengthen connections.

Language Translation and Real-Time Translation

The reason for these technological advancements is the research on voice processing and machine learning, and companies will utilize the development of voice processing technology to play a larger role in language processing. Imagine being able to understand every language on Earth through a simple wearable device. Thanks to advancements in voice technology, this could become a reality.

References Author: Zhihu User

Reference Author: Pineapple Boy

Link: https://www.zhihu.com/question/36983728/answer/131423105

Source: Zhihu

Some content adapted from NetEase Smart Studio (WeChat official account smartman163)