Papers/2609.12085
🧪 Test?View on arXiv

What Counts as a Mistake? Annotating Recitation Events in Quran Memorization Transcripts

Not provided in the content

annotationASRerror detectionmachine learning
2609.12085
Builder Relevance
60%
2h ago

Abstract

The paper discusses the challenges of distinguishing various types of mistakes in Quran recitation transcripts and presents a human annotation study with an executable evaluator.

Reality Card

Core Claim

The study achieved a label-aware F1 score of 0.525 and localization F1 score of 0.826, demonstrating the effectiveness of the annotation process for identifying mistakes in Quran recitation.

Method / Result

The best-performing model achieved a label-aware F1 score ranging from 0.143 to 0.892 across different runs.

Limitations

The main limitation is that no run was annotated before building, which affects the reproducibility of the results.

Paper to code

Verified implementation resources so builders can test the paper’s claims instead of stopping at the abstract.

No verified implementation link has been attached yet. AIBuzzHub will keep this panel separate from unverified search results.
← Back to all papers