Indic Voice Technologies for an Inclusive Digital India: Toolkit for Developers
Contains acronyms p. 7
"This toolkit sets out a layered, lifecycle-oriented approach to building inclusive and robust speech artificial intelligence (AI) systems. It brings together strategies for diverse and representative data collection, linguistically informed model training, rigorous quality control, and deployment optimisation under real-world constraints, alongside embedded Responsible AI (RAI) practices." (Executive summary, page 10)
1 Building Voice Technology for India: Issues and Best Practices, 15
2. Ensuring Diverse Representation, 18
2.1 Challenges in Ensuring Diverse Representation, 19
2.2 Best Practices for Ensuring Diverse Representation, 23
3. Enhancing Data Quality and Building Inclusive Applications, 30
3.1 Challenges in Ensuring Data Quality and Building Inclusive Applications, 31
3.2 Best Practices for Quality and Building Inclusive Applications, 34
4. Embedding Responsible AI (RAI) Practices, 40
4.1 Challenges in Operationalising RAI Practices, 41
4.2 Best Practices for Operationalising RAI Practices, 42
Appendix 1 Current Status in Open Source Voice Technologies in India, 52
2. Ensuring Diverse Representation, 18
2.1 Challenges in Ensuring Diverse Representation, 19
2.2 Best Practices for Ensuring Diverse Representation, 23
3. Enhancing Data Quality and Building Inclusive Applications, 30
3.1 Challenges in Ensuring Data Quality and Building Inclusive Applications, 31
3.2 Best Practices for Quality and Building Inclusive Applications, 34
4. Embedding Responsible AI (RAI) Practices, 40
4.1 Challenges in Operationalising RAI Practices, 41
4.2 Best Practices for Operationalising RAI Practices, 42
Appendix 1 Current Status in Open Source Voice Technologies in India, 52
