Update README.md

README.md

@@ -137,17 +137,16 @@ timing, transcription, overlap ratio, source utterance id).
 
 - **Audio:** 16 kHz, mono, 16-bit PCM WAV
 - **Mixtures:** 37,800 total — **30,000** train / **3,900** dev / **3,900** test
-- **Speaker counts:** 2spk, 3spk, 4spk (equal share per split)
+- **Speaker counts:** 2spk, 3spk, 4spk
 - **Total duration:** ≈ 82.5 h (train ≈ 65.4 h, dev ≈ 8.5 h, test ≈ 8.6 h)
 - **Languages:** English (`ENG`) + Korean (`KOR`), including bilingual mixtures
-- **Speakers:** 514 unique (322 EN + 192 KO), **speaker-disjoint** across splits
+- **Speakers:** 514 speakers
 
 ---
 
 ## Reliability & Verification
 
-This dataset is designed for **trustworthy evaluation**. The following properties are
-programmatically verified by `dataset_audio.py` (raw log: `dataset_audio.0905.log`).
+This dataset is designed for **trustworthy evaluation**.
 
 ### 1. Speaker-disjoint splits
 
@@ -178,13 +177,13 @@ distribution.
 
 Overlap is not concentrated at one difficulty. Mixtures span the full range from
 clean (no overlap) to fully overlapped, with the harder end growing as speaker count
-increases (2→4 spk), which is the expected and desired behaviour.
+increases (2-4 spk), which is the expected and desired behaviour.
 
 <img src="./figure/overlap_distribution.png" width="70%">
 
 ### 4. Complete, machine-readable metadata
 
-Every mixture has a paired JSON with per-speaker ground truth (see schema below),
+Every mixture has a paired JSON with per-speaker ground truth,
 enabling exact, reproducible analysis of language / gender / overlap / timing.
 
 ---
@@ -198,7 +197,7 @@ KO-only mixtures remain represented at every count:
 
 Across the 514 unique speakers, the male / female balance is:
 
-<img src="./figure/gender_ratio.png" width="70%">
+<img src="./figure/gender_ratio.png" width="60%">
 
 ---
 
@@ -207,11 +206,11 @@ Across the 514 unique speakers, the male / female balance is:
 ```
 Librittsr-KOR-n/
 ├── train/ {2spk,3spk,4spk}/
-│   ├── raw/   <mixture_id>.wav      # 16 kHz mono mixture
-│   └── json/  <mixture_id>.json     # per-mixture ground-truth metadata
+│   ├── raw/   <mixture_id>.wav
+│   └── json/  <mixture_id>.json
 ├── dev/   {2spk,3spk,4spk}/ {raw,json}/
 ├── test/  {2spk,3spk,4spk}/ {raw,json}/
-└── scp/   {2spk,3spk,4spk}/         # Kaldi-style file lists
+└── scp/   {2spk,3spk,4spk}/
 ```
 
 **Filename convention.** A mixture filename is the concatenation of its source
@@ -273,7 +272,7 @@ root = "path/to/Librittsr-KOR-n"
 split, spk = "test", "2spk"
 
 for wav in sorted(glob.glob(f"{root}/{split}/{spk}/raw/*.wav")):
-    audio, sr = sf.read(wav)                     # 16 kHz mono mixture
+    audio, sr = sf.read(wav) 
     meta = json.load(open(wav.replace("/raw/", "/json/").replace(".wav", ".json")))
     n_src   = meta["speaker_count"]
     overlap = meta["overlapped_ratio"]
@@ -288,8 +287,7 @@ for wav in sorted(glob.glob(f"{root}/{split}/{spk}/raw/*.wav")):
 This dataset is derived from multiple sources with **different licenses**:
 
 - **English audio** originates from **LibriTTS-R** (CC BY 4.0).
-- **Korean audio** originates from a Korean read-speech corpus — **confirm and state
-  its license before public release.**
+- **Korean audio** originates from a Multi-speaker Speech Synthesis Data by National Information Society Agency (NIA)
 
 Because of these mixed terms the top-level tag is set to `other`. Please verify the
 Korean-source license and update this section (and the `license`/`license_name` fields)