Rapid Growth in Voice Recognition AI Market... Related Services Also Advancing

[Asia Economy Reporter Donghyun Choi] As the voice recognition artificial intelligence (AI) market experiences rapid growth, related services are attracting attention by playing active roles across various industries through technological advancements.

According to the industry on the 12th, VITO, a call application developed by AI startup ReturnZero that allows users to "see" calls, converts call recordings into text and displays them like a messenger. In addition to the basic function of separating the voices of both parties within seconds after the call ends and showing them like a messenger, users can easily search for desired content. Without listening to the entire call recording from start to finish, users can simply click on the desired part for easy partial playback. Additional services such as call content editing, sharing, and note-taking are also supported, making it a convenient app for everyday use.

ReturnZero introduced the 'Call Preview' feature in September last year through continuous technological advancement. This feature allows users to preview previous call contents with the caller regardless of whether the contact is saved, enabling preparation before answering the phone.

ReturnZero provides its service through its proprietary STT (Speech To Text) technology-based 'Somers Engine,' which converts call audio into text, and the 'Moses Engine,' which analyzes users' voices to separate speakers. VITO has gained popularity among a wide range of users, from the general public to professionals whose work requires frequent calls, such as salespeople, lawyers, real estate agents, and insurance planners, by offering high accuracy and high-quality features.

Google has been leading the global translation market since introducing its AI-based neural machine translation (GNMT) service in November 2016.

The neural network technology applied by Google evolved from translating words and phrases within sentences to enabling sentence-level translation similar to human language use. This technology has been applied to eight languages, including English, French, German, Spanish, Portuguese, Chinese, Japanese, and Turkish. Currently, Google supports translation services for a total of 103 languages. It is expected that more refined translation quality will be achievable in the future as more data is learned through machine learning.

With the COVID-19 pandemic accelerating remote work and non-face-to-face operations, English-speaking remote business has also increased. Consequently, the need for conversational skills has grown, leading to the emergence of ‘AI speaking tutor apps’ in the English education sector.

‘Speak,’ developed by SpeakEasy Labs, is an AI-based English speaking education app service. It helps learners study alone anytime and anywhere they want, overcoming spatial and temporal constraints of learning. It also complements aspects of autonomous learning that traditional methods could not address.

Speak allows learners to repeatedly practice English expressions used in daily life with AI, naturally increasing the number of speaking attempts. It also provides missions where learners complete expressions by filling in blanks within sentences. The Speak app offers over 1,500 conversation practice contents and presents various situations that can occur in daily life, making it highly practical.

In March last year, KB Kookmin Bank introduced an AI banker capable of real-time two-way communication. After its debut at the AI Experience Zone in the KB Kookmin Bank headquarters in Yeouido, the AI banker’s response speed and performance have been further improved, and content has been added as it is sequentially expanded to the Yeouido branch, Insight branch, and Donam-dong branch.

The AI banker derives optimal answers to customer inquiries through KB Kookmin Bank’s proprietary financial-specialized language model KB-STA. Currently applied to kiosks, it plays various roles such as providing instructions on using intelligent automated machines (STM), automated teller machines (ATM), introducing financial products, and guiding necessary documents. To this end, the AI banker has learned about 800 financial terms and 1,300 scenarios.

KB Kookmin Bank’s AI banker enables safe service for customers who prefer face-to-face transactions amid the COVID-19 situation. It is also expected to effectively reduce customers’ perceived waiting times.