Reconstructed Slides: Why your agents need decision traces, not just documents — Zach Blumenfeld, Neo4j

Source Video

Why your agents need decision traces, not just documents — Zach Blumenfeld, Neo4j

Method

This deck is reconstructed from the existing video frame captures by detecting likely slide regions with OpenCV, cropping/upscaling those regions, deduplicating similar crops, and OCRing the cropped slide images locally. It is a cleaner companion to the full-stage frame deck.

Reconstructed Slides

• Source frame: slide-001.jpg
• Crop: full [0, 0, 960, 540] score 157.8

• Source frame: slide-002.jpg
• Crop: full [0, 0, 960, 540] score 166.23

• Source frame: slide-003.jpg
• Crop: contour [0, 0, 960, 540] score 177.64

• Source frame: slide-004.jpg
• Crop: contour [0, 0, 960, 540] score 164.49

• Source frame: slide-005.jpg
• Crop: contour [0, 0, 960, 540] score 165.93

• Source frame: slide-006.jpg
• Crop: contour [0, 0, 960, 540] score 176.62

• Source frame: slide-007.jpg
• Crop: contour [0, 0, 960, 540] score 177.69

• Source frame: slide-008.jpg
• Crop: full [0, 0, 960, 540] score 176.89

• Source frame: slide-009.jpg
• Crop: contour [0, 0, 960, 540] score 177.18

• Source frame: slide-010.jpg
• Crop: full [0, 0, 960, 540] score 167.87

• Source frame: slide-011.jpg
• Crop: full [0, 0, 960, 540] score 177.22

• Source frame: slide-012.jpg
• Crop: full [0, 0, 960, 540] score 166.36

• Source frame: slide-013.jpg
• Crop: full [0, 0, 960, 540] score 164.33

• Source frame: slide-014.jpg
• Crop: full [0, 0, 960, 540] score 167.19

• Source frame: slide-015.jpg
• Crop: full [0, 0, 960, 540] score 176.8

• Source frame: slide-016.jpg
• Crop: full [0, 0, 960, 540] score 166.79

Dense Scene-Detected Slide Candidates

• youtube B9h9ovW5H9U dense slides