File size: 13,590 Bytes
cb39c05
 
 
 
 
 
 
 
 
 
 
 
 
 
ba92724
cb39c05
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
95e1515
cb39c05
 
 
 
 
 
 
 
 
 
95e1515
cb39c05
 
 
 
 
 
 
 
 
 
95e1515
cb39c05
 
 
 
 
 
 
 
 
 
95e1515
cb39c05
 
 
 
 
 
 
 
 
 
 
 
 
 
 
95e1515
 
 
 
 
 
cb39c05
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
95e1515
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
cb39c05
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
95e1515
cb39c05
 
 
 
 
 
 
 
 
 
 
1a6b8d0
cb39c05
1a6b8d0
 
95e1515
cb39c05
 
 
95e1515
 
 
 
 
 
 
 
 
 
 
cb39c05
95e1515
cb39c05
 
 
95e1515
 
cb39c05
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
"""
Gradio tab for speaker separation workflow.

Provides UI for separating speakers from multi-speaker audio files.
"""

import json
import logging
import tempfile
from pathlib import Path
from typing import List, Optional, Tuple

import gradio as gr

from src.services.speaker_separation import SpeakerSeparationService

logger = logging.getLogger(__name__)


def create_speaker_separation_tab() -> gr.Tab:
    """
    Create the speaker separation tab for the Gradio interface.

    Returns:
        Configured Gradio Tab component
    """
    with gr.Tab("Speaker Separation") as tab:
        gr.Markdown(
            """
            # πŸ‘₯ Speaker Separation

            Analyze multi-speaker audio files to automatically detect and separate
            individual speakers into separate audio streams.

            Upload an audio file with multiple speakers, and this tool will:
            - Detect all speakers automatically
            - Separate each speaker's audio
            - Export clean individual streams
            """
        )

        with gr.Row():
            with gr.Column(scale=1):
                # Input Section
                gr.Markdown("### πŸ“€ Input File")

                input_audio = gr.Audio(
                    label="Multi-Speaker Audio File",
                    type="filepath",
                    sources=["upload"],
                )

                # Configuration Section
                gr.Markdown("### βš™οΈ Speaker Detection Settings")

                with gr.Row():
                    min_speakers = gr.Slider(
                        minimum=1,
                        maximum=10,
                        value=2,
                        step=1,
                        label="Minimum Speakers",
                        info="Minimum number of speakers expected",
                    )

                    max_speakers = gr.Slider(
                        minimum=1,
                        maximum=10,
                        value=5,
                        step=1,
                        label="Maximum Speakers",
                        info="Maximum number of speakers expected",
                    )

                num_speakers = gr.Slider(
                    minimum=0,
                    maximum=10,
                    value=0,
                    step=1,
                    label="Exact Speaker Count (0 = auto-detect)",
                    info="Set to non-zero to specify exact number",
                )

                with gr.Accordion("Output Settings", open=True):
                    output_format = gr.Radio(
                        choices=["m4a", "wav", "mp3"],
                        value="m4a",
                        label="Output Format",
                    )

                    with gr.Row():
                        sample_rate = gr.Slider(
                            minimum=8000,
                            maximum=48000,
                            value=44100,
                            step=100,
                            label="Sample Rate (Hz)",
                        )

                        bitrate = gr.Dropdown(
                            choices=["128k", "192k", "256k", "320k"],
                            value="192k",
                            label="Bitrate",
                        )

                # Action Buttons
                with gr.Row():
                    separate_btn = gr.Button("πŸš€ Separate Speakers", variant="primary", size="lg")
                    clear_btn = gr.ClearButton(components=[input_audio], value="πŸ—‘οΈ Clear")

            with gr.Column(scale=1):
                # Output Section
                gr.Markdown("### πŸ“Š Results")

                # Status
                status_output = gr.Textbox(
                    label="Status",
                    placeholder="Ready to process...",
                    interactive=False,
                    lines=3,
                )

                # Progress indicator (will be updated during processing)
                progress_bar = gr.Progress()

                # Results summary
                summary_output = gr.JSON(
                    label="Separation Summary",
                    visible=False,
                )

                # Speaker details
                with gr.Accordion("Speaker Details", open=True, visible=False) as details_accordion:
                    speaker_table = gr.Dataframe(
                        headers=["Speaker", "Duration (s)", "Confidence"],
                        label="Detected Speakers",
                        interactive=False,
                    )

                # Download Section
                gr.Markdown("### πŸ’Ύ Downloads")

                output_files = gr.File(
                    label="Separated Speaker Files",
                    file_count="multiple",
                    interactive=False,
                    visible=False,
                )

                report_file = gr.File(
                    label="Separation Report (JSON)",
                    interactive=False,
                    visible=False,
                )

        # Examples and Tips
        gr.Markdown("### πŸ“š Usage Tips")
        gr.Markdown(
            """
            **How to Use:**

            1. **Upload Audio**: Select an M4A, WAV, or MP3 file with multiple speakers
            2. **Configure Detection**:
               - Use min/max speakers for auto-detection (recommended)
               - Or set exact speaker count if you know it
            3. **Choose Output**: Select format, sample rate, and bitrate
            4. **Separate**: Click the button and wait for processing
            5. **Download**: Get individual speaker files and a detailed report

            **Best Practices:**
            - Clear audio with distinct speakers works best
            - If you know the exact speaker count, specify it for better results
            - Processing time scales with file duration (expect ~2x realtime)
            - M4A format provides best quality-to-size ratio
            - For long files (>1 hour), expect several minutes of processing

            **Troubleshooting:**
            - If fewer speakers detected than expected, try increasing max_speakers
            - If too many speakers detected, try increasing min_speakers
            - For overlapping speech, the tool will assign to the dominant speaker
            """
        )

        # Event Handler
        separate_btn.click(
            fn=_separate_speakers_handler,
            inputs=[
                input_audio,
                min_speakers,
                max_speakers,
                num_speakers,
                output_format,
                sample_rate,
                bitrate,
            ],
            outputs=[
                status_output,
                summary_output,
                speaker_table,
                output_files,
                report_file,
                details_accordion,
            ],
        )

    return tab


def _separate_speakers_handler(
    input_audio: Optional[str],
    min_speakers: int,
    max_speakers: int,
    num_speakers: int,
    output_format: str,
    sample_rate: int,
    bitrate: str,
    progress=gr.Progress(),
) -> Tuple[str, dict, list, list, str, gr.Accordion]:
    """
    Handler function for speaker separation.

    Args:
        input_audio: Path to input audio file
        min_speakers: Minimum speakers to detect
        max_speakers: Maximum speakers to detect
        num_speakers: Exact speaker count (0 = auto)
        output_format: Output format (m4a, wav, mp3)
        sample_rate: Output sample rate
        bitrate: Output bitrate
        progress: Gradio progress tracker

    Returns:
        Tuple of (status, summary, speaker_data, output_files, report_file, accordion_visibility)
    """
    try:
        # Validate inputs
        if not input_audio:
            return (
                "❌ Error: Please upload an audio file",
                {},
                [],
                [],
                None,
                gr.update(visible=False),
            )

        input_path = Path(input_audio)
        if not input_path.exists():
            return (
                f"❌ Error: File not found: {input_audio}",
                {},
                [],
                [],
                None,
                gr.update(visible=False),
            )

        # Validate speaker counts
        if min_speakers > max_speakers and num_speakers == 0:
            return (
                f"❌ Error: Minimum speakers ({min_speakers}) cannot exceed maximum ({max_speakers})",
                {},
                [],
                [],
                None,
                gr.update(visible=False),
            )

        # Validate sample rate for M4A
        if output_format == "m4a" and sample_rate > 48000:
            return (
                f"❌ Error: Sample rate {sample_rate} exceeds M4A limit of 48000 Hz",
                {},
                [],
                [],
                None,
                gr.update(visible=False),
            )

        # Use exact speaker count if specified
        if num_speakers > 0:
            min_speakers = num_speakers
            max_speakers = num_speakers

        # Create temporary output directory
        output_dir = Path(tempfile.mkdtemp(prefix="speaker_separation_"))

        # Initialize service
        progress(0.1, desc="Initializing speaker separation models...")
        service = SpeakerSeparationService()

        # Progress callback
        def progress_callback(stage: str, current: float, total: float):
            # Interpret float-based (0.0-1.0) vs integer-based formats
            if total == 1.0:
                pct = 0.1 + (current * 0.8)  # Scale float 0.0-1.0 to 10-90%
            else:
                pct = 0.1 + (current / total) * 0.8  # Scale integer to 10-90%
            progress(pct, desc=stage)

        progress(0.1, desc="Starting speaker separation...")

        # Run separation
        report = service.separate_and_export(
            input_file=str(input_path),
            output_dir=str(output_dir),
            min_speakers=min_speakers,
            max_speakers=max_speakers,
            output_format=output_format,
            sample_rate=sample_rate,
            bitrate=bitrate,
            progress_callback=progress_callback,
        )

        # Check if result is an error report
        if report.get("status") == "failed":
            error_message = f"❌ **Error ({report['error_type']}):** {report['error']}"
            # Save error report
            error_report_path = output_dir / "error_report.json"
            with open(error_report_path, "w") as f:
                json.dump(report, f, indent=2)
            return (
                error_message,
                {},
                [],
                [],
                str(error_report_path),
                gr.update(visible=False),
            )

        progress(0.9, desc="Preparing results...")

        # Build speaker table data
        speaker_data = []
        for output_info in report["output_files"]:
            speaker_data.append(
                [
                    output_info["speaker_id"],
                    f"{output_info['duration']:.1f}",
                    f"{output_info.get('confidence', 1.0):.2f}",
                ]
            )

        # Collect output files
        output_file_paths = [
            str(output_dir / output_info["file"]) for output_info in report["output_files"]
        ]

        # Save report to file
        report_path = output_dir / "separation_report.json"
        with open(report_path, "w") as f:
            json.dump(report, f, indent=2)

        # Build status message
        status = f"""βœ… Separation Complete!

🎀 Detected {report["speakers_detected"]} speaker(s)
⏱️ Processed in {report["processing_time_seconds"]:.1f} seconds
πŸ“ Output saved to temporary directory

You can download the separated audio files and detailed report below.
"""

        # Build summary
        summary = {
            "speakers_detected": report["speakers_detected"],
            "processing_time": f"{report['processing_time_seconds']:.1f}s",
            "input_duration": f"{report['input_duration_seconds']:.1f}s",
            "output_format": output_format,
            "sample_rate": f"{sample_rate} Hz",
        }

        if "overlapping_segments" in report:
            summary["overlapping_segments"] = report["overlapping_segments"]

        progress(1.0, desc="Done!")

        return (
            status,
            gr.JSON(value=summary, visible=True),
            speaker_data,
            gr.File(value=output_file_paths, visible=True),
            gr.File(value=str(report_path), visible=True),
            gr.update(visible=True),
        )

    except Exception as e:
        # Catch any unexpected errors not handled by the service
        logger.exception("Unexpected error in speaker separation")
        error_report = {
            "status": "failed",
            "error": f"Unexpected error: {str(e)}",
            "error_type": "processing",
        }
        # Save error report
        error_report_path = output_dir / "error_report.json"
        with open(error_report_path, "w") as f:
            json.dump(error_report, f, indent=2)
        return (
            f"❌ **Error:** {error_report['error']}",
            {},
            [],
            [],
            str(error_report_path),
            gr.update(visible=False),
        )