KUDO AI Gets an 84% Increase in Accuracy Translating Names, Products, and Places

(Monday, 26 June 2023. New York City, NY) – Continuing their quality-first approach to speech translation, the KUDO team rolls out another round of upgrades to their AI Speech Translator—this time, focusing on the accuracy of names. With this upgrade, precision in transcription and translation of names can expect to increase by up to 84%, making KUDO AI (patent pending) a more robust and reliable solution for global events, conferences, and product launches.

In today’s rapidly evolving tech environment, speech recognition and speech translation systems, play an increasingly vital role. Their accuracy and precision is constantly improving. Yet, these systems often face significant challenges when it comes to the accurate processing of names. These may include people’s names, product names, organizations, places, acronyms unique to an industry or an organization, and other unique identifiers. 

There are several factors that make proper names difficult for machines, including (but not limited to):

1. Variety and Uniqueness: The wide range of possible names across the world makes it extremely challenging to build a system that can accurately recognize all of them.
2. Accents and Pronunciation Variations: Even common names can be pronounced differently based on the speaker’s accent, dialect, or personal preference. For example, the name “Andrea” can be pronounced as “AN-dree-uh”, “an-DREY-uh”, or even “ahn-DRAY-uh” based on the speaker’s background.
3. Phonetic Similarities: Some names sound very similar to other words or names, making them difficult for speech recognition systems to differentiate. For example, the name “Neil” could easily be mistaken for “kneel” or “Neal.”
4. Spelling Variations: Different spellings can represent the same sound (for example, “Sean”, “Shawn”, and “Shaun”), which adds an extra level of complexity for these systems.
5. Uncommon and Foreign Names: Speech recognition systems are usually trained on a dataset of common words and names. If a name is rare, foreign, or spelled in an unconventional way, the system might not recognize it properly. 

KUDO AI‘s latest upgrade is poised to improve the way speech systems handle proper names by fine-tuning the speech recognition engine on the fly. Before the start of a session, clients can pass a custom list of relevant names and phrases, increasing their importance, and thus making it more likely to be recognized by the engine. 

Our internal test reports that with this new feature, precision in transcription and translation of proper names has seen an increase of up to 84%. This significant leap forward in technology is beneficial for events where specific names are frequently used—imagine a global product launch where accurate recognition and translation of the product name is crucial for seamless communication. 

By enabling users to tailor the translation model to their specific needs, we are ensuring that our technology is adaptable, user-centric, and capable of delivering unparalleled accuracy in transcribing and translating proper names. As we continue to innovate, we are excited to see how this unique feature of KUDO AI will open new avenues for communication, collaboration, and understanding in an increasingly interconnected world. 

About KUDO

KUDO is the world leader in providing real-time multilingual solutions that enable people to communicate effortlessly in any language⁠—on any platform. Their network of 12,000 professional language interpreters, combined with their ground-breaking Speech Translator, empower organizations of all sizes to collaborate more efficiently, with greater inclusivity, and on an international scale. KUDO Inc. is a New-York based technology start-up founded and managed by language and conferencing industry insiders seeking to create a world in which everyone has the power to understand and be understood in their own language. More info at www.kudoway.com.

Contact

Chloë Walford
chloe@kudoway.com