Persian Language Added to Speechmatics: A New Milestone
Speechmatics adds Persian, exploring writing and speech variances, achieving rapid language model success in just weeks for diverse customer use.
File
Understanding Persian - bringing a new language to Speechmatics
Added on 01/29/2025
Speakers
add Add new speaker

Speaker 1: We've recently added our 49th language, Persian. Adding a new language starts with understanding both how the language is spoken, but also how it's written. Something interesting with Persian is that it can be written in multiple alphabets. There are different ways of writing it from different regions around the world and different ways of speaking it. So we had to figure out how do you transcribe both informal and formal speech from people in various regions that may be speaking the same language, but expecting it to be written differently. Strangely enough, Speechmatics learning a new language is much like any of us learning a language on a language course. We start off by immersing ourselves in how the language is written and spoken. And like any good teacher, we have something to guide us along the way. We then take throughout the term, maybe a few small tests to kind of guide our learning. Then it's time for the final exam. For us, the final exam is a test set or some data that we haven't seen before. And we transcribe it with a reference transcript that we know is perfect. And we compare the two. And we publish this as our word error rate. Speechmatics learns a new language much more quickly than we do as humans. The whole process took us three to four weeks from starting to understand the Persian language to releasing our models to customers. Handing it over to our customers was a delight. And seeing them use our tech with the new language that was really beneficial to them was perfect.

ai AI Insights
Summary

Generate a brief summary highlighting the main points of the transcript.

Generate
Title

Generate a concise and relevant title for the transcript based on the main themes and content discussed.

Generate
Keywords

Identify and highlight the key words or phrases most relevant to the content of the transcript.

Generate
Enter your query
Sentiments

Analyze the emotional tone of the transcript to determine whether the sentiment is positive, negative, or neutral.

Generate
Quizzes

Create interactive quizzes based on the content of the transcript to test comprehension or engage users.

Generate
{{ secondsToHumanTime(time) }}
Back
Forward
{{ Math.round(speed * 100) / 100 }}x
{{ secondsToHumanTime(duration) }}
close
New speaker
Add speaker
close
Edit speaker
Save changes
close
Share Transcript