Dialpad India Successfully Trains Six Indian Languages as Part of MUCS 2021 ASR Challenge

Dialpad plans to grow the India software engineering team by up to 70% over the next 12 months to focus heavily on research and development

Dialpad Inc., the industry leader in AI-powered communication and collaboration, announced its participation in the MUCS 2021: MUltilingual and Code-Switching ASR Challenges for Low Resource Indian Languages. The Dialpad India team outperformed the program baselines and successfully trained multilingual automatic speech recognition (ASR) models in six Indian languages — Gujarati, Hindi, Marathi, Oriya, Tamil and Telugu. Dialpad is making its data preparation recipe, pre-trained models and Recurrent Neural Network Language Modeling research for all six languages available through open-source to advance the development of ASR.

Marketing Technology News: Alliance Develops Shared Diversity and Inclusion Goals

“We thank the MUCS 2021 organizers for making this project possible and hope our contribution to the Indian language ASR community will be useful for further research.”

Multilinguality for natural language processing (NLP) is a rising challenge in the ASR community, especially for low-resource languages in emerging markets, and area of active research in both academia and commercial industries. By exploring challenges and solutions in these areas, Dialpad is strengthening its artificial intelligence (AI) and Voice Intelligence (Vi™) capabilities to address the challenges of multilinguality and low-resource languages.

Marketing Technology News: MarTech Interview with Clarence Kam, Chief Product Officer at JW Player

“Through this exercise, we identified challenges to multilingual ASR methods and were able to develop and propose several potential solutions. Data from this project will also be analyzed and leveraged internally to improve Dialpad Voice Intelligence,” said Amit Kaul, Vice President of Engineering and India Country Manager, Dialpad. “We thank the MUCS 2021 organizers for making this project possible and hope our contribution to the Indian language ASR community will be useful for further research.”

Dialpad continues to grow its India software engineering team with up to 50 new hires over the next 12 months. The India team will focus on Dialpad conferencing capabilities, product integrations and cloud platform infrastructure to support artificial intelligence (AI), a critical component as Dialpad expands research and development efforts for its proprietary Voice Intelligence technology. Vi delivers real-time business optimization, including call coaching, automated note-taking, sentiment tracking and transcription analysis to businesses of all sizes.