Evaluating Tibetan ASR With Segmented Word Error Rate: Beyond Character-Level Metrics
Previous work on Tibetan automatic speech recognition (ASR) has relied on character error rate (CER) or syllable error rate (SER), since the absence of explicit word boundaries in Tibetan orthography complicates the calculation of word error rate (WER). In this paper, we instantiate Segmented Word E...
Sparad:
| Huvudupphov: | , , |
|---|---|
| Materialtyp: | Artigo |
| Språk: | Inglês |
| Utgiven: |
IEEE
2026-01-01
|
| Serie: | IEEE Access |
| Ämnen: | |
| Länkar: | https://ieeexplore.ieee.org/document/11592371/ |
| Taggar: |
Inga taggar, Lägg till första taggen!
|