Article complete

Soket AI Labs Unveils DHRITH, India’s Emotion-Aware Speech Recognition System

DHRITH includes features such as emotion tagging, code-mixed fluency, speaker diarisation, and context-aware transcription.

A speaker in a blue suit presents on stage, with a blue background featuring the Soket logo and diagrams.

Key takeaways

  • Soket AI Labs launches DHRITH, India's first emotion-aware speech recognition system.
  • DHRITH captures tone, emotion, and multilingual speech patterns unique to Indian users.
  • The system includes features like emotion tagging and context-aware transcription for accurate communication.
  • Developed under Project EKΛ, DHRITH aims to enhance India's AI potential through audio data.
  • Future support includes an API and resources for developers to integrate DHRITH into applications.

Takeaways drafted with AI from the article text.

The system was created to address the gap in India’s ASR technology landscape. “In India, most data doesn’t live in text – it lives in voice,” Soket AI Labs said in its post on LinkedIn. 

The firm explained that existing tools could not “capture tone, emotion, code switching and speaker context with precision.”

DHRITH was developed under Project EKΛ, an open-source foundation model supported by IndiaAI and the electronics and IT ministry. The project focuses on unlocking India’s AI potential through audio data that is “rich, expressive, and multilingual.”

Soket AI Labs stated that this ASR system includes features such as emotion tagging, code-mixed fluency, speaker diarisation, and context-aware transcription. These functions enable it to process how Indians speak across languages and situations.

DHRITH has become the company’s in-house ASR “designed to listen the way India speaks.” It is expected to support research, enterprise applications, and eventually developers through an upcoming API and technical blog.

The system is also designed to “power the next generation of Indian AI” by integrating cultural and linguistic nuances that are often overlooked by conventional systems. A demonstration video and technical details are set to be released soon.

Published

Spotted an error? Tell our editors. Read our editorial standards and corrections.

More in this story

Soket AI Labs · 5 stories
  1. Soket AI Labs Unveils Pragna-1B, Multilingual Indic Language Model
  2. Soket AI Labs Partners with Google Cloud to Boost Pragna-1B Model
  3. Soket AI Labs Launches Realtime Speech API, Offers Multilingual Capabilities
  4. MeitY to Back Sarvam AI, Soket AI Labs, Gnani.ai for IndiaAI Mission
  5. You are here: Soket AI Labs Unveils DHRITH, India’s Emotion-Aware Speech Recognition System
About the author

Technology Journalist

Covering deep and emerging tech: semiconductors, quantum, robotics, space tech, drones and GCCs. Connect via socials below or email:

All stories
Belamy · Daily

The day’s AI news, before your first meeting.