Merge pull request #2429 from jackykang7777/patch-2

Update release-notes-stt.md for Semantic Segmentation Feature Release

Author: prmerger-automator[bot]

articles/ai-services/speech-service/includes/release-notes/release-notes-stt.md

@@ -7,6 +7,32 @@ ms.author: eur
 
 ### January 2025 release
 
+#### New Feature - Semantic Segmentation
+Announcing the release of a new feature: Semantic Segmentation. This feature integrates a punctuation module inside decoder that segments audio based on semantic information, resulting in more logical and precise segmentation boundaries.
+Key Benefits:
+- Improved Segmentation Accuracy: By leveraging semantic information, this feature significantly reduces instances of long segments caused by the absence of pauses in the input audio.
+- Reduce latency caused by under-segmentation: The overall latency for the Speech-to-Text service is reduced, with a 40%-60% reduction in the length of the longest 5% of segments.
+- Over-Segmentation Mitigation: This feature also helps prevent over-segmentation by delaying segmentation when a better sentence can be formed.
+
+Supported Locales:
+
+<li>English (en-US, en-GB)</li>
+<li>Chinese (zh-CN, zh-HK)</li>
+<li>Japanese (ja-JP)</li>
+<li>Korean (ko-KR)</li>
+<li>German (de-DE)</li>
+<li>French (fr-FR)</li>
+<li>Italian (it-IT)</li>
+<li>Spanish (es-ES, es-MX)</li>
+<li>Hindi (hi-IN)</li>
+<li>Portuguese (pt-BR, pt-PT)</li>
+<li>Turkish (tr-TR)</li>
+<li>Russian (ru-RU)</li>
+<li>Thai (th-TH)</li>
+<li>Indonesian (id-ID)</li>
+
+For implementation details, please refer to the documentation: [How to Recognize Speech](https://learn.microsoft.com/azure/ai-services/speech-service/how-to-recognize-speech?pivots=programming-language-csharp) in section "Semantic Segmentation".
+
 #### Real-time speech to text - New English model release
 
 Announcing the release of the latest English speech model (en-US, en-CA), which brings substantial improvements across various performance metrics. Below are the key highlights of this release: