Context-Aware ASR for Mandarin Technical Lectures
2026-07-06 • Sound
Sound
AI summaryⓘ
The authors studied how well speech recognition systems understand English technical terms mixed into Mandarin lectures. They found that traditional error rates don’t show when these important terms are missed. To fix this, they created a new way to test and improve recognizing technical terms by letting the system first guess common terms and then use that guess to improve recognition. This approach helped the system catch more technical terms without making more mistakes overall. Their work shows that using the lecture’s own context helps find key terms better than usual methods.
Mandarin ASRCharacter Error RateTechnical-term recognitionTwo-pass decodingGlossary promptingSpeech recognitionMachine learning lecturesTerm recallTerm precisionContextual language modeling
Authors
Ho-Lam Chung, Yiming Chen, Hung-yi Lee
Abstract
Technical lectures mix Mandarin speech with English technical terms. These terms carry the core meaning of the lecture, yet they occupy few characters. Character error rate (CER) therefore hides their recognition failures. We study whether lecture context helps recognize these terms. We build a term-rich Mandarin AI/ML lecture benchmark, and we define term-centric metrics that measure technical-term recognition directly. We then propose a two-pass, reference-free decoding method. The first pass runs segment-only ASR. We extract the most frequent technical terms from the first-pass hypotheses, and we prompt the recognizer with this self-built glossary in the second pass. Across five ASR backbones, the first-pass glossary raises term recall for every model and holds or lowers CER on all five. On Breeze-ASR-25 it lifts term recall from 52.50% to 60.13% while lowering CER, and a hybrid that adds a small external term list reaches 62.05% recall and 82.73% term precision. Lecture context, recovered from the model's own output, is a practical signal for technical-term recognition. Term-centric evaluation exposes errors that CER misses.