Update README.md

README.md

@@ -53,7 +53,7 @@ license: apache-2.0
 
 
 ### Metadata Format
-We store all audio metadata in a standardized JSON format, where the core fields include `utt_id` (unique identifier for each audio segment), `rover_result` (ROVER result of three ASR transcriptions), `confidence` (confidence score of text transcription), `jyutping_confidence` (confidence score of Cantonese pinyin transcriptions), and `duration` (audio duration); speaker attributes include `speaker_id`, `gender`, and `age`; audio quality assessment metrics include `sample_rate`, `DNSMOS`, and `SNR`; timestamp information includes `timestamp` (precisely recording segment boundaries with `start` and `end`); and extended metadata under the `meta_info` field includes `program` (program name), `region` (geographical information), `link` (original content link), and `domain` (domain classification).
+We store all audio metadata in a standardized JSON format, where the core fields include `utt_id` (unique identifier for each audio segment), `rover_result` (ROVER result of three ASR transcriptions), `confidence` (confidence score of text transcription), and `duration` (audio duration); speaker attributes include `gender`, and `age`; audio quality assessment metrics include `sample_rate`, `DNSMOS`; timestamp information includes `timestamp` (precisely recording segment boundaries with `start` and `end`); and extended metadata under the `meta_info` field includes `program` (program name), `region` (geographical information), `link` (original content link), and `domain` (domain classification).
 
 
 #### 📂 Content Tree