azcomputerguru/claudetools
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
c760e430c08e477c5dbffcb28fbc93d58549566b
claudetools/projects/radio-show/audio-processor/test-data/transcripts/2016-s8e43/diarization.json
Mike Swanson c760e430c0 radio: bumper detection in diarizer + full archive download script 
Adds a transcript-driven bumper filter to the diarization pipeline. When
a transcript segment matches qa_extractor's promo/bumper signatures, the
overlapping audio windows are labeled BUMPER and the WavLM cosine match
is skipped. Prevents music/promo from being matched against speaker
profiles (the failure mode Mike caught in 2018-s10e18 @ 09:20-10:05).

Code changes:
- src/voice_profiler.py: identify_speakers() takes optional skip_ranges
  parameter; windows whose midpoint falls in a skip range get labeled
  "[bumper]" and skip cosine match
- src/diarizer.py: diarize() takes optional transcript_path; pre-computes
  bumper time ranges via qa_extractor._is_promo_or_bumper, passes to
  identify_speakers; adds BUMPER speaker label
- benchmark.py: passes transcript_path to diarize()

Aggregate impact across 9-episode test set:
  Tara attribution: 4880s -> 3680s  (-1200s / -25%)
  Q&A pairs: 17 -> 19 (+2)
    (bumper-flagged segments had been disrupting conversation detection
     in 2017-s9e30 and 2018-s10e18)
  CALLER total: 1320s -> 1190s  (bumpers previously labeled CALLER moved)
  Per-episode bumpers caught: 1-8, total ~165 bumper segments across set

Remaining Tara false positives are real callers acoustically similar to
Tara (Christopher in 2018, Kay in 2012, William and Charles in 2015) and
guest Clay in 2015-s7e19 — those need profile rebuild + Clay profile,
not bumper filtering.

Adds download_full_archive.py — resumable mirror-style downloader that
walks IX server's /home/gurushow/public_html/archive/{year}/ and copies
all MP3s to archive-data/episodes/. Run is in progress (~589 files,
~10-15GB). Used to source clean profile windows for the remaining
co-hosts (Tara rebuild, Clay, Tony, Rob, Randall, producers).

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-04-27 16:17:50 -07:00

839 lines
16 KiB
JSON
Raw Blame History

{
"num_speakers": 4,
"speaker_map": {
"HOST": "HOST",
"CALLER": "CALLER",
"BUMPER": "BUMPER",
"CO-HOST": "CO-HOST"
},
"turns": [
{
"speaker": "BUMPER",
"start": 0.0,
"end": 35.0,
"confidence": 1.0
},
{
"speaker": "CO-HOST",
"start": 30.0,
"end": 40.0,
"confidence": 0.96
},
{
"speaker": "HOST",
"start": 35.0,
"end": 55.0,
"confidence": 0.97
},
{
"speaker": "CO-HOST",
"start": 50.0,
"end": 60.0,
"confidence": 0.96
},
{
"speaker": "HOST",
"start": 55.0,
"end": 100.0,
"confidence": 0.97
},
{
"speaker": "CO-HOST",
"start": 95.0,
"end": 120.0,
"confidence": 0.9
},
{
"speaker": "HOST",
"start": 115.0,
"end": 140.0,
"confidence": 0.94
},
{
"speaker": "CO-HOST",
"start": 135.0,
"end": 160.0,
"confidence": 0.97
},
{
"speaker": "HOST",
"start": 155.0,
"end": 175.0,
"confidence": 0.96
},
{
"speaker": "CALLER",
"start": 170.0,
"end": 185.0,
"confidence": 0.77
},
{
"speaker": "HOST",
"start": 180.0,
"end": 275.0,
"confidence": 0.87
},
{
"speaker": "CALLER",
"start": 270.0,
"end": 295.0,
"confidence": 0.82
},
{
"speaker": "HOST",
"start": 290.0,
"end": 345.0,
"confidence": 0.97
},
{
"speaker": "CALLER",
"start": 340.0,
"end": 355.0,
"confidence": 0.81
},
{
"speaker": "HOST",
"start": 350.0,
"end": 370.0,
"confidence": 0.96
},
{
"speaker": "CO-HOST",
"start": 365.0,
"end": 380.0,
"confidence": 0.94
},
{
"speaker": "HOST",
"start": 375.0,
"end": 525.0,
"confidence": 0.97
},
{
"speaker": "CO-HOST",
"start": 520.0,
"end": 535.0,
"confidence": 0.99
},
{
"speaker": "HOST",
"start": 530.0,
"end": 550.0,
"confidence": 0.97
},
{
"speaker": "CO-HOST",
"start": 545.0,
"end": 555.0,
"confidence": 0.98
},
{
"speaker": "HOST",
"start": 550.0,
"end": 565.0,
"confidence": 0.98
},
{
"speaker": "BUMPER",
"start": 560.0,
"end": 595.0,
"confidence": 1.0
},
{
"speaker": "CO-HOST",
"start": 590.0,
"end": 600.0,
"confidence": 0.93
},
{
"speaker": "CALLER",
"start": 595.0,
"end": 605.0,
"confidence": 0.84
},
{
"speaker": "HOST",
"start": 600.0,
"end": 1055.0,
"confidence": 0.96
},
{
"speaker": "CO-HOST",
"start": 1050.0,
"end": 1060.0,
"confidence": 0.93
},
{
"speaker": "HOST",
"start": 1055.0,
"end": 1160.0,
"confidence": 0.99
},
{
"speaker": "BUMPER",
"start": 1155.0,
"end": 1205.0,
"confidence": 1.0
},
{
"speaker": "CO-HOST",
"start": 1200.0,
"end": 1215.0,
"confidence": 0.98
},
{
"speaker": "CALLER",
"start": 1210.0,
"end": 1220.0,
"confidence": 0.8
},
{
"speaker": "CO-HOST",
"start": 1215.0,
"end": 1235.0,
"confidence": 0.86
},
{
"speaker": "HOST",
"start": 1230.0,
"end": 1295.0,
"confidence": 0.96
},
{
"speaker": "CO-HOST",
"start": 1290.0,
"end": 1300.0,
"confidence": 0.95
},
{
"speaker": "HOST",
"start": 1295.0,
"end": 1335.0,
"confidence": 0.98
},
{
"speaker": "CO-HOST",
"start": 1330.0,
"end": 1345.0,
"confidence": 0.94
},
{
"speaker": "HOST",
"start": 1340.0,
"end": 1380.0,
"confidence": 0.97
},
{
"speaker": "CO-HOST",
"start": 1375.0,
"end": 1395.0,
"confidence": 0.98
},
{
"speaker": "HOST",
"start": 1390.0,
"end": 1410.0,
"confidence": 0.97
},
{
"speaker": "CO-HOST",
"start": 1405.0,
"end": 1415.0,
"confidence": 0.98
},
{
"speaker": "HOST",
"start": 1410.0,
"end": 1795.0,
"confidence": 0.96
},
{
"speaker": "CO-HOST",
"start": 1790.0,
"end": 1875.0,
"confidence": 0.99
},
{
"speaker": "HOST",
"start": 1870.0,
"end": 1950.0,
"confidence": 0.98
},
{
"speaker": "CO-HOST",
"start": 1945.0,
"end": 1960.0,
"confidence": 0.98
},
{
"speaker": "HOST",
"start": 1955.0,
"end": 2025.0,
"confidence": 0.98
},
{
"speaker": "CO-HOST",
"start": 2020.0,
"end": 2030.0,
"confidence": 0.92
},
{
"speaker": "BUMPER",
"start": 2025.0,
"end": 2060.0,
"confidence": 1.0
},
{
"speaker": "HOST",
"start": 2055.0,
"end": 2105.0,
"confidence": 0.98
},
{
"speaker": "CO-HOST",
"start": 2100.0,
"end": 2115.0,
"confidence": 0.98
},
{
"speaker": "HOST",
"start": 2110.0,
"end": 2145.0,
"confidence": 0.98
},
{
"speaker": "CO-HOST",
"start": 2140.0,
"end": 2150.0,
"confidence": 0.97
},
{
"speaker": "HOST",
"start": 2145.0,
"end": 2235.0,
"confidence": 0.96
},
{
"speaker": "CO-HOST",
"start": 2230.0,
"end": 2240.0,
"confidence": 0.93
},
{
"speaker": "HOST",
"start": 2235.0,
"end": 2270.0,
"confidence": 0.98
},
{
"speaker": "CO-HOST",
"start": 2265.0,
"end": 2280.0,
"confidence": 0.95
},
{
"speaker": "HOST",
"start": 2275.0,
"end": 2285.0,
"confidence": 0.97
},
{
"speaker": "CO-HOST",
"start": 2280.0,
"end": 2290.0,
"confidence": 0.92
},
{
"speaker": "HOST",
"start": 2285.0,
"end": 2340.0,
"confidence": 0.89
},
{
"speaker": "CO-HOST",
"start": 2335.0,
"end": 2360.0,
"confidence": 0.94
},
{
"speaker": "HOST",
"start": 2355.0,
"end": 2375.0,
"confidence": 0.94
},
{
"speaker": "CO-HOST",
"start": 2370.0,
"end": 2410.0,
"confidence": 0.98
},
{
"speaker": "HOST",
"start": 2405.0,
"end": 2415.0,
"confidence": 0.93
},
{
"speaker": "CO-HOST",
"start": 2410.0,
"end": 2425.0,
"confidence": 0.98
},
{
"speaker": "HOST",
"start": 2420.0,
"end": 2545.0,
"confidence": 0.96
},
{
"speaker": "CO-HOST",
"start": 2540.0,
"end": 2550.0,
"confidence": 0.97
},
{
"speaker": "HOST",
"start": 2545.0,
"end": 2610.0,
"confidence": 0.93
},
{
"speaker": "CO-HOST",
"start": 2605.0,
"end": 2620.0,
"confidence": 0.93
},
{
"speaker": "HOST",
"start": 2615.0,
"end": 2680.0,
"confidence": 0.98
},
{
"speaker": "CO-HOST",
"start": 2675.0,
"end": 2685.0,
"confidence": 0.9
},
{
"speaker": "HOST",
"start": 2680.0,
"end": 2690.0,
"confidence": 0.97
},
{
"speaker": "CO-HOST",
"start": 2685.0,
"end": 2770.0,
"confidence": 0.95
},
{
"speaker": "HOST",
"start": 2765.0,
"end": 2795.0,
"confidence": 0.9
},
{
"speaker": "CO-HOST",
"start": 2790.0,
"end": 2820.0,
"confidence": 0.98
},
{
"speaker": "HOST",
"start": 2815.0,
"end": 2855.0,
"confidence": 0.91
},
{
"speaker": "CO-HOST",
"start": 2850.0,
"end": 2860.0,
"confidence": 0.98
},
{
"speaker": "HOST",
"start": 2855.0,
"end": 2865.0,
"confidence": 0.97
},
{
"speaker": "CO-HOST",
"start": 2860.0,
"end": 2885.0,
"confidence": 0.91
},
{
"speaker": "HOST",
"start": 2880.0,
"end": 2905.0,
"confidence": 0.95
},
{
"speaker": "CO-HOST",
"start": 2900.0,
"end": 2935.0,
"confidence": 0.98
},
{
"speaker": "HOST",
"start": 2930.0,
"end": 2995.0,
"confidence": 0.98
},
{
"speaker": "CO-HOST",
"start": 2990.0,
"end": 3010.0,
"confidence": 0.97
},
{
"speaker": "HOST",
"start": 3005.0,
"end": 3035.0,
"confidence": 0.95
},
{
"speaker": "CO-HOST",
"start": 3030.0,
"end": 3060.0,
"confidence": 0.97
},
{
"speaker": "HOST",
"start": 3055.0,
"end": 3140.0,
"confidence": 0.95
},
{
"speaker": "CO-HOST",
"start": 3135.0,
"end": 3150.0,
"confidence": 0.98
},
{
"speaker": "HOST",
"start": 3145.0,
"end": 3205.0,
"confidence": 0.97
},
{
"speaker": "CO-HOST",
"start": 3200.0,
"end": 3210.0,
"confidence": 0.89
},
{
"speaker": "HOST",
"start": 3205.0,
"end": 3220.0,
"confidence": 0.96
},
{
"speaker": "CO-HOST",
"start": 3215.0,
"end": 3225.0,
"confidence": 0.96
},
{
"speaker": "HOST",
"start": 3220.0,
"end": 3230.0,
"confidence": 0.96
},
{
"speaker": "CO-HOST",
"start": 3225.0,
"end": 3260.0,
"confidence": 0.91
},
{
"speaker": "HOST",
"start": 3255.0,
"end": 3270.0,
"confidence": 0.98
},
{
"speaker": "CO-HOST",
"start": 3265.0,
"end": 3275.0,
"confidence": 0.97
},
{
"speaker": "HOST",
"start": 3270.0,
"end": 3350.0,
"confidence": 0.96
},
{
"speaker": "CO-HOST",
"start": 3345.0,
"end": 3375.0,
"confidence": 0.98
},
{
"speaker": "HOST",
"start": 3370.0,
"end": 3390.0,
"confidence": 0.94
},
{
"speaker": "BUMPER",
"start": 3385.0,
"end": 3435.0,
"confidence": 1.0
},
{
"speaker": "HOST",
"start": 3430.0,
"end": 3970.0,
"confidence": 0.98
},
{
"speaker": "BUMPER",
"start": 3965.0,
"end": 4020.0,
"confidence": 1.0
},
{
"speaker": "CO-HOST",
"start": 4015.0,
"end": 4025.0,
"confidence": 0.95
},
{
"speaker": "CALLER",
"start": 4020.0,
"end": 4030.0,
"confidence": 0.85
},
{
"speaker": "HOST",
"start": 4025.0,
"end": 4050.0,
"confidence": 0.92
},
{
"speaker": "CO-HOST",
"start": 4045.0,
"end": 4055.0,
"confidence": 0.93
},
{
"speaker": "HOST",
"start": 4050.0,
"end": 4095.0,
"confidence": 0.87
},
{
"speaker": "CO-HOST",
"start": 4090.0,
"end": 4100.0,
"confidence": 0.92
},
{
"speaker": "HOST",
"start": 4095.0,
"end": 4190.0,
"confidence": 0.98
},
{
"speaker": "CO-HOST",
"start": 4185.0,
"end": 4200.0,
"confidence": 0.93
},
{
"speaker": "HOST",
"start": 4195.0,
"end": 4215.0,
"confidence": 0.98
},
{
"speaker": "CO-HOST",
"start": 4210.0,
"end": 4225.0,
"confidence": 0.98
},
{
"speaker": "HOST",
"start": 4220.0,
"end": 4240.0,
"confidence": 0.97
},
{
"speaker": "CO-HOST",
"start": 4235.0,
"end": 4250.0,
"confidence": 0.99
},
{
"speaker": "HOST",
"start": 4245.0,
"end": 4385.0,
"confidence": 0.98
},
{
"speaker": "CO-HOST",
"start": 4380.0,
"end": 4400.0,
"confidence": 0.99
},
{
"speaker": "HOST",
"start": 4395.0,
"end": 4425.0,
"confidence": 0.96
},
{
"speaker": "CO-HOST",
"start": 4420.0,
"end": 4435.0,
"confidence": 0.99
},
{
"speaker": "HOST",
"start": 4430.0,
"end": 4440.0,
"confidence": 0.95
},
{
"speaker": "CO-HOST",
"start": 4435.0,
"end": 4445.0,
"confidence": 0.91
},
{
"speaker": "HOST",
"start": 4440.0,
"end": 4460.0,
"confidence": 0.98
},
{
"speaker": "CO-HOST",
"start": 4455.0,
"end": 4465.0,
"confidence": 0.94
},
{
"speaker": "HOST",
"start": 4460.0,
"end": 4515.0,
"confidence": 0.94
},
{
"speaker": "CO-HOST",
"start": 4510.0,
"end": 4525.0,
"confidence": 0.97
},
{
"speaker": "HOST",
"start": 4520.0,
"end": 4570.0,
"confidence": 0.96
},
{
"speaker": "CO-HOST",
"start": 4565.0,
"end": 4580.0,
"confidence": 0.94
},
{
"speaker": "HOST",
"start": 4575.0,
"end": 4655.0,
"confidence": 0.97
},
{
"speaker": "BUMPER",
"start": 4650.0,
"end": 4695.0,
"confidence": 1.0
},
{
"speaker": "CO-HOST",
"start": 4690.0,
"end": 4715.0,
"confidence": 0.94
},
{
"speaker": "HOST",
"start": 4710.0,
"end": 4790.0,
"confidence": 0.97
},
{
"speaker": "CALLER",
"start": 4785.0,
"end": 4795.0,
"confidence": 0.83
},
{
"speaker": "HOST",
"start": 4790.0,
"end": 4805.0,
"confidence": 0.89
},
{
"speaker": "CALLER",
"start": 4800.0,
"end": 4815.0,
"confidence": 0.79
},
{
"speaker": "HOST",
"start": 4810.0,
"end": 4935.0,
"confidence": 0.92
},
{
"speaker": "CALLER",
"start": 4930.0,
"end": 4940.0,
"confidence": 0.82
},
{
"speaker": "HOST",
"start": 4935.0,
"end": 4945.0,
"confidence": 0.86
},
{
"speaker": "CALLER",
"start": 4940.0,
"end": 4950.0,
"confidence": 0.82
},
{
"speaker": "HOST",
"start": 4945.0,
"end": 5080.0,
"confidence": 0.86
},
{
"speaker": "CALLER",
"start": 5075.0,
"end": 5085.0,
"confidence": 0.85
},
{
"speaker": "HOST",
"start": 5080.0,
"end": 5220.0,
"confidence": 0.86
},
{
"speaker": "CO-HOST",
"start": 5215.0,
"end": 5225.0,
"confidence": 0.98
},
{
"speaker": "HOST",
"start": 5220.0,
"end": 5325.0,
"confidence": 0.95
}
]
}
Reference in New Issue View Git Blame Copy Permalink