Preprint
Factors affecting ASR performance: A study using state of the art ASR models in Indic Languages
Agneedh Basu et al., arXiv preprint, 2026
Read the originalRead on October 17, 2026
- What it is
- A study of the output of several open-source speech recognition models on Indian speech datasets in five languages. It tests how speaker traits and recording conditions change word error rate, and for Hindi it analyzes telephone codecs, bit depth, resampling and background noise.
- How much weight it carries
- A 2026 preprint that does not yet appear to be peer-reviewed. It reports correlations, not causes, and this guide cites only what the study analyzes, not its conclusions about specific systems.
Facts taken from this source
- A 2026 study of open-source speech recognition models on Indian speech analyzes Hindi, Bengali, Kannada, Telugu and Marathi.
“We analyze linguistic, speaker-level, and acoustic factors across Hindi, Bengali, Kannada, Telugu, and Marathi.” (Abstract)