#natural language process...
Mar 2026
PARSA-Bench: A Comprehensive Persian Audio-Language Model Benchmark
Across most tasks, text-only baselines outperform their audio counterparts, so audio understanding rather than language knowledge remains the main limitation, and supplying the transcript alongside the audio lifts weak models to near their text-only level.
Mohammad Javad Ranjbar Kalahroodi, M. Amini, Parmis Bathayan et al.
· arXiv.org · 2 citations