Plot added

.gitignore

@@ -1 +1,4 @@
-__pycache__/*
+__pycache__/*
+flagged/*
+input.*
+output.*

README.md

@@ -1,12 +1,12 @@
 ---
-title: Choral Quartets To MIDI
+title: Choral Quartets F0 Extractor
 emoji: 🐠
 colorFrom: yellow
 colorTo: blue
 sdk: gradio
 sdk_version: 3.38.0
 app_file: app.py
-pinned: false
+pinned: true
 license: cc
 ---
 

app.py

@@ -2,8 +2,9 @@ import gradio as gr
 from cqfe_utils import cqfe
 cqfe_interface = gr.Interface(fn=cqfe,
                               inputs=gr.Audio(type='filepath', format='wav', label='Audio Input File'),
-                              outputs=gr.File(type='file', label='F0 Output Files'),
-                              title="Choral Quartets F0 Extractor",
+                              outputs=[gr.Plot(label='F0 Estimation Plot'),
+                                        gr.File(type='file', label='F0 Output Files')],
+                              title="Choral Quartets F0 Extractor (v0.1.2)",
                               description="An application that uses Multi-Pitch Estimation and Voice Assignment to transform audio files with Choral Quartets recordings into files (CSV, HDF5 and MIDI) containing F0 estimations for each voice (Soprano, Alto, Tenor and Bass). The processing may take a few minutes.")
 
 cqfe_interface.launch()

app_test.ipynb

@@ -2,24 +2,76 @@
  "cells": [
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 2,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Running on local URL:  http://127.0.0.1:7860\n",
+      "\n",
+      "To create a public link, set `share=True` in `launch()`.\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div><iframe src=\"http://127.0.0.1:7860/\" width=\"100%\" height=\"500\" allow=\"autoplay; camera; microphone; clipboard-read; clipboard-write;\" frameborder=\"0\" allowfullscreen></iframe></div>"
+      ],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/plain": []
+     },
+     "execution_count": 2,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "2023-07-21 21:02:27.573436: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:432] Loaded cuDNN version 8600\n",
+      "2023-07-21 21:02:28.575934: I tensorflow/tsl/platform/default/subprocess.cc:304] Start cannot spawn child process: No such file or directory\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "1/1 [==============================] - 10s 10s/step\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "2023-07-21 21:02:35.800368: I tensorflow/compiler/xla/stream_executor/cuda/cuda_blas.cc:606] TensorFloat-32 will be used for the matrix multiplication. This will only be logged once.\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "1/1 [==============================] - 0s 369ms/step\n"
+     ]
+    }
+   ],
    "source": [
     "import gradio as gr\n",
-    "from cqfe_utils import cqfe\n"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": [
+    "from cqfe_utils import cqfe\n",
     "cqfe_interface = gr.Interface(fn=cqfe,\n",
     "                              inputs=gr.Audio(type='filepath', format='wav', label='Audio Input File'),\n",
-    "                              outputs=gr.File(type='file', label='F0 Output Files'),\n",
-    "                              title=\"Choral Quartets F0 Extractor\",\n",
+    "                              outputs=[gr.Plot(label='F0 Estimation Plot'),\n",
+    "                                        gr.File(type='file', label='F0 Output Files')],\n",
+    "                              title=\"Choral Quartets F0 Extractor (v0.1.2)\",\n",
     "                              description=\"An application that uses Multi-Pitch Estimation and Voice Assignment to transform audio files with Choral Quartets recordings into files (CSV, HDF5 and MIDI) containing F0 estimations for each voice (Soprano, Alto, Tenor and Bass). The processing may take a few minutes.\")\n",
     "\n",
     "cqfe_interface.launch()"

cqfe_utils.py

@@ -145,10 +145,8 @@ def song_to_midi(sop, alto, ten, bass):
 
 ############################################################
 
-def song_to_tables(sop, alto, ten, bass):
+def song_to_dataframe(sop, alto, ten, bass):
 
-    savepath_csv = './output.csv'
-    savepath_hdf5 = './output.hdf5'
     timescale = np.arange(0, 0.011609977 * (sop.shape[1]), 0.011609977)[:sop.shape[1]]
 
     s_argmax = vec_bin_to_freq(np.argmax(sop, axis=0))
@@ -160,10 +158,8 @@ def song_to_tables(sop, alto, ten, bass):
     columns = ['Timestep', 'Soprano', 'Alto', 'Tenor', 'Bass']
 
     df = pd.DataFrame(data, columns=columns)
-    df.to_csv(savepath_csv, mode='w', header=True)
-    df.to_hdf(savepath_hdf5, key='F0', mode='w', complevel=9, complib='blosc', append=False, format='table')
 
-    return savepath_csv, savepath_hdf5
+    return df
 
 ############################################################
 
@@ -284,10 +280,27 @@ def get_va_prediction(model, f0_matrix):
 ############################################################
 
 def cqfe(audiofile, mpe=late_deep_cnn_model(), va=mask_voas_cnn_v2_model()):
+    
+    savepath_csv = './output.csv'
+    savepath_hdf5 = './output.hdf5'
+
     mpe_pred = get_mpe_prediction(mpe, audiofile)
     s_pred, a_pred, t_pred, b_pred = get_va_prediction(va, mpe_pred)
+
     output_midi = song_to_midi(s_pred, a_pred, t_pred, b_pred)
-    output_csv, output_hdf5 = song_to_tables(s_pred, a_pred, t_pred, b_pred)
-    return [output_midi, output_csv, output_hdf5]
+
+    output_df = song_to_dataframe(s_pred, a_pred, t_pred, b_pred)
+    output_df.to_csv(savepath_csv, mode='w', header=True)
+    output_df.to_hdf(savepath_hdf5, key='F0', mode='w', complevel=9, complib='blosc', append=False, format='table')
+    ax1 = output_df.plot.scatter(x='Timestep', y='Bass', s=1, color='#2f29e3', label='Bass')
+    ax2 = output_df.plot.scatter(x='Timestep', y='Tenor', s=1, color='#e36129', label='Tenor', ax=ax1)
+    ax3 = output_df.plot.scatter(x='Timestep', y='Alto', s=1, color='#29e35a', label='Alto', ax=ax1)
+    ax4 = output_df.plot.scatter(x='Timestep', y='Soprano', s=1, color='#d3d921', label='Soprano', ax=ax1)
+    ax1.set_xlabel('Time (s)')
+    ax1.set_ylabel('Freq (Hz)')
+    fig = ax1.get_figure()
+    fig.set_dpi(200)
+
+    return fig, [output_midi, savepath_csv, savepath_hdf5]
 
 ############################################################

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "Choral-Quartets-F0-Extractor"
-version = "0.1.1"
+version = "0.1.2"
 description = "An application that uses Multi-Pitch Estimation and Voice Assignment to transform audio files with Choral Quartets recordings into files (CSV, HDF5 and MIDI) containing F0 estimations for each voice (Soprano, Alto, Tenor and Bass)."
 authors = ["André Paiva (Xornotor) <[email protected]>"]
 license = "cc"
@@ -21,6 +21,7 @@ hdf5plugin
 h5py
 tables
 librosa
+matplotlib
 
 
 [build-system]

requirements.txt

@@ -9,4 +9,5 @@ pandas
 hdf5plugin
 h5py
 tables
-librosa
+librosa
+matplotlib