Anton A Smirnov
Papers
1
Total Citations
11
H-Index
1
About
Anton A. Smirnov is a researcher in speech synthesis and artificial intelligence, with a focus on developing neural network architectures that enable more natural and adaptable voice generation. His key research areas include speaker- and style-adaptive speech synthesis, neural network embeddings, and data-efficient training methods for text-to-speech systems. Smirnov’s most notable contribution is his work on speaker/style-dependent neural network speech synthesis, where he introduced a novel architecture that leverages speaker and style embeddings to generate synthesized speech in a specific voice and speaking style using only a small amount of target training data. This approach, detailed in his highly cited 2020 paper (11 citations), addresses a critical challenge in speech synthesis: achieving high-quality personalization with minimal data. By mapping discrete speaker and style variables into continuous embedding spaces, Smirnov’s method enables more flexible and efficient voice cloning, with applications in virtual assistants, accessibility tools, and entertainment. His work has been recognized for its practical impact, offering a scalable solution for generating diverse, natural-sounding speech while reducing the computational and data requirements traditionally associated with neural network-based synthesis.
Research Focus
Key Achievements
Top Papers
- 1