Markdown source

Slides: Beyond Transcription: Building Voice AI That Understands Conversations — Hervé Bredin, pyannoteAI

Source Video

Beyond Transcription: Building Voice AI That Understands Conversations — Hervé Bredin, pyannoteAI

Relationship To World's Fair 2026

These slides are extracted from a public AI Engineer YouTube video connected to World's Fair 2026. Speaker-matched clips are supporting context unless later confirmed as exact session recordings; official livestream recordings are day-level/event-level source material.

Related Scheduled Sessions

• No individual scheduled session mapping has been assigned yet; treat this as an event livestream deck.

Extracted Slides

• Recreated text/layout view: open HTML recreation
• AI slide classifier: content_slide confidence 0.98
• Text source: advanced OCR rapidocr-live/border-trim/contrast.
• OCR decision: ready — dense slide with screenshot, diagram, and small text

Slide text:

Transcription

WHAT was said

AIE

lorem ipsum dolor sit amet consectetur adipiscing elit

hendrerit morbi sed nunc morbi convallis erat non

mollis ex sollicitudin porttitor velit at

pyannoteAl

Google DeepMind

AlEnginee

• Recreated text/layout view: open HTML recreation
• AI slide classifier: content_slide confidence 0.99
• Text source: agent_vision.
• OCR decision: ready — dense slide with code snippet, chart, and small labels

Slide text:

What can go wrong?

hopefully the answer is not "the live demo"

• Recreated text/layout view: open HTML recreation
• AI slide classifier: content_slide confidence 0.99
• Text source: advanced OCR rapidocr-live/bright-screen/contrast.
• OCR decision: ready — dense chart slide with small labels and benchmark text

Slide text:

State of the art (Q4 2025)

Conversational telephone speech (CTS)

DIHARDCTS/61conversations/2speakers FoseAorrn KieCeienCersen

AIE pyannoteAi

X

271

151 X.OA

22.5

KN

5.5A Ln

HipLsreney mengMini

BPGE

pyannote.ai/benchmark

AI Engineer

EUROPE

• Recreated text/layout view: open HTML recreation
• AI slide classifier: content_slide confidence 0.98
• Text source: advanced OCR rapidocr-live/full.
• OCR decision: ready — dense comparison slide with chart and screenshot text

Slide text:

What can go wrong? theblamegame Wirnport the b.Th Op Aye,o hat isnot it Eit Oseck Pe a the bet Deh afe evarivated.

Subetution Deletionnsertion OM

AIE tcorc-WER 50% mytahealth/Crispeisper Coherelabs/cohere-transcribe-03-2026 ihe-ganite/ganite-4.0-llsoeoch nodel 5.52 5.42 6.67 280.02 84.05 RTFx 524.88 LS6 Ogen 8,44

ihm-ganit/granite-sech-3.3-2 Open 8,9

ieganite/gnite-sch-3.3-8 5.74 145.42 Open 8.98

301 a.5 0n/03-AS-1.78 vidia/canaty-o 0-2.5 5.63 147.93 418.28 Open Open 95'01

20% s7 nicrosoft/Phi-4-multinodal-festruct usefelse $.66 20`9 151.1 448.15 Open Open 11.09 10.68

6.05 3386.02 Open

evidla/oarakeet-tet-0.6b-3 6.32 3332.74 ope 11.39

6,42 146.23 Open

Whisperx Parckeet

pyannote.ai/blog/stt-orchestration hf.co/spaces/hf-audio/open_asr_leaderboard

AIEngineer

EUROPE

Hidden Non-Slide Evidence

• `slide-001.jpg` — speaker_stage confidence 0.99; speaker at podium with projected slide; not a readable slide capture
• `slide-006.jpg` — sponsor_logo confidence 0.99; logo/end card, not a content slide

Classification audit: raw/sources/slide-ai-classification/slides/mFLlVpnGpds/audit.json

Slide-Derived Subjects To Review

Subject extraction uses video title, related session titles/descriptions, transcript context, and OCR text when available. OCR is best-effort and should be reviewed against the embedded slide images.