As a pioneer and innovator in cross-language communication solutions, Timekettle has relentlessly advanced the field and introduced actual AI translation devices that provided the company with significant experience from actual usage. Initially offering handheld translators for simple exchanges, Timekettle's breakthrough came with the WT2 Plus Translator Earbuds, featuring HybridComm 1.0 for seamless, one-way translation and a speaker mode for broader communication. The subsequent launch of the WT2 Edge Real-time Translator Earbuds, with HybridComm 2.0, enabled natural, bidirectional conversations. The latest advancement, powered by HybridComm 3.0 features AI Semantic Segmentation, which the company says significantly enhances translation efficiency and accuracy, marking a new era in near-human-level AI translation technology.
Powered by Timekettle’s latest TurboFast Technology under HybridComm 3.0, the new AI Semantic Segmentation technology ensures a faster translation experience, helping the translated speech align closer to the speaker’s rhythm, and shortening the delayed of translated output, which is crucial communication scenarios. The technology begins by grasping the core meaning of sentences. Once it aligns closely with the "entirely correct" benchmark, it begins translating while still processing the original voice, allowing users to hear the translated audio sooner.
In well-established simultaneous translation solutions, as used in international conferences and official government affairs, human translators have to listen to each sentence and translate the correct meaning to a microphone, while continuing to listen to the rest of the speech. For the receivers, this means a need to abstract from the original speech and focus only on the audio that is received in-ear, which obviously is disconnected from the actual discourse, gestures and lip movements. When using an automatic language translation solution with earbuds, that delay was typically longer (requiring external processing), which forced the users to detach completely from live interactions and experience. Those delays can now be shortened with local processing, while accuracy improves with AI technology.
According to Timekettle, thanks to the AI Semantic Segmentation, the delay in receiving translated audio has been reduced by 200%, bringing its real-time translation capabilities on par with live interpretation. Internal test conducted by the company recently revealed that with the AI Semantic Segmentation, a translated speech only delays a sentence lag time, while other traditional speech translation devices may experience significant delays of 30 to 40 seconds - which is usable for asking an occasional question, but makes a conversation impossible.
“This significant reduction, cutting over 20 seconds of waiting time, has revolutionized continuous communication. Picture being at a conference room where attendees can instantly hear the translation of a sentence right as the speaker starts the next. Such a seamless flow of communication marks a new era of cross-language interaction, paving the way for a more connected world,” Timekettle states.
The AI Semantic Segmentation technology is now available as an update for the Timekettle WT2 Edge Real-time Translator Earbuds, M3 Language Translator Earbuds, and the latest X1 AI Interpreter Hub solution (One-on-one mode, Listen & Play mode, and Ask & Go mode for now; Voice Call mode and more will be available from June 28, 2024).
Established in 2016, Timekettle is dedicated to advancing cross-language communication and its products and solutions have been tested in practice by more than 400,000 users who live abroad, during traveling, and for multi-language meetings and classrooms. Particularly for visitors to China, Timekettle has provided a practical solution.
At CES 2024, the company introduced the X1 AI Interpreter Hub, a standalone device that comes with earbuds, enabling users to initiate conversations swiftly without additional apps or complex setups. For the first time, the X1 device also includes an innovative one dial feature, which enables real-time cross-language calls.
The X1 can be used as a handheld translator, supporting translation for meetings of up to 20 people, and up to five languages.
More recently, Timekettle launched its AI Translation Lab app platform to more rapidly introduce cutting-edge new features and gather immediate user feedback to avoid releasing underdeveloped functionalities. Understanding the importance of accurate translations for proper nouns and specialized terms in everyday conversations, the AI Translation Lab features a Custom Lexicon, which allows users to create their own terms linked to specific translations, ensuring precision from these personalized glossaries.
The software also introduced AI Language Tutor with a focus on enhancing user’s fluency and pronunciation of foreign languages. At present, Timekettle's AI Translation Lab supports Chinese, English, and Spanish languages but more languages will be added in upcoming updates.
www.timekettle.co
