samples

.gitignore

@@ -0,0 +1,19 @@
+gradio_env/
+.DS_Store
+.env
+.venv/
+# Exclude large audio files (use parquet instead)
+*.wav
+*.mp3
+*.flac
+common_samples/
+organized_samples/
+
+# Keep parquet files (small, no Git LFS needed)
+!parquet_samples/
+!*.parquet
+
+# Python
+__pycache__/
+*.py[cod]
+*$py.class

211_audio_samples.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b4a2376819631b88b8ed6bdcae272acece1d0a0f6a962be531b077632d239732
+size 33996737

4014_audio_samples.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f1cf39ba7ea2c895fe485db1235bfffe58b48f150b79834c2eb919deec3e2eb7
+size 35781880

F02_audio_samples.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:051fed6bac89a4f6031970afdedf4cb17f15c1ad9c557efb562a7152045c7806
+size 15836445

F04_audio_samples.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:02339455600a256f8a2d35379546747a5edfc7e02dd92a35ebaf057b996780e2
+size 16499443

FC02_audio_samples.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2abedd7a17821e894236af63b47291ee31877ab7a6df61628291c7b221e663aa
+size 11918414

M02_audio_samples.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:34d2884c27120fea60df11e5507ecd14373eecfdb4bb3d77d66636eaf7765eec
+size 18661702

M04_audio_samples.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f54f7c7d0b22bbe39bbe70b4337dbf58480adc46b59cf1a0e063f5b2def08efe
+size 9262819

MC01_audio_samples.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:10b513c89b82b57a06b3206741092c7e72a7851550cc841971266803a61e529a
+size 14827983

parquet_audio_index.json

@@ -0,0 +1,21 @@
+{
+  "format": "parquet_audio",
+  "description": "Audio samples stored as parquet files with audio arrays",
+  "sample_rate": 16000,
+  "total_speakers": 8,
+  "files": [
+    "211_audio_samples.parquet",
+    "F04_audio_samples.parquet",
+    "FC02_audio_samples.parquet",
+    "F02_audio_samples.parquet",
+    "MC01_audio_samples.parquet",
+    "M04_audio_samples.parquet",
+    "M02_audio_samples.parquet",
+    "4014_audio_samples.parquet"
+  ],
+  "usage": {
+    "load_audio": "json.loads(df[\"audio_array\"].iloc[0])",
+    "get_sample_rate": "df[\"sample_rate\"].iloc[0]",
+    "play_in_gradio": "gr.Audio((sample_rate, np.array(audio_array)))"
+  }
+}

requirements.txt

@@ -0,0 +1,45 @@
+# Pathological Speech Synthesis Comparison App
+# Python dependencies for the interactive demonstration
+
+# Web interface
+gradio>=4.0.0
+
+# Data manipulation
+pandas>=1.5.0
+numpy>=1.21.0
+
+# Parquet file support
+pyarrow>=10.0.0
+
+# Audio processing (used in conversion scripts)
+librosa>=0.9.0
+soundfile>=0.12.0
+
+# Utilities
+pathlib2>=2.3.7; python_version < '3.4'
+
+# Optional: For enhanced visualizations
+matplotlib>=3.5.0
+seaborn>=0.11.0
+
+# Optional: For audio analysis
+scipy>=1.9.0
+
+# JSON handling (built-in, but listing for completeness)
+# json - built-in module
+
+# File operations (built-in)
+# shutil - built-in module
+# os - built-in module
+
+# Web server dependencies (usually installed with gradio)
+uvicorn>=0.18.0
+fastapi>=0.85.0
+
+# Additional gradio dependencies
+pillow>=9.0.0
+requests>=2.25.0
+
+# For reproducibility
+wheel>=0.37.0
+setuptools>=65.0.0