Reconstructed Slides: Cognitive Shield Real Time Real Smart - Rachna Srivastava

Source Video

Cognitive Shield Real Time Real Smart - Rachna Srivastava

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, 623] score 171.48

• Source frame: slide-002.jpg
• Crop: full [0, 0, 960, 623] score 171.89

• Source frame: slide-003.jpg
• Crop: full [0, 0, 960, 623] score 172.1

• Source frame: slide-004.jpg
• Crop: full [0, 0, 960, 623] score 171.3

• Source frame: slide-005.jpg
• Crop: full [0, 0, 960, 623] score 168.84

• Source frame: slide-006.jpg
• Crop: full [0, 0, 960, 623] score 167.52

• Source frame: slide-007.jpg
• Crop: full [0, 0, 960, 623] score 168.28

• Source frame: slide-008.jpg
• Crop: full [0, 0, 960, 623] score 170.87

• Source frame: slide-009.jpg
• Crop: full [0, 0, 960, 623] score 171.09

• Source frame: slide-010.jpg
• Crop: full [0, 0, 960, 623] score 170.92

• Source frame: slide-011.jpg
• Crop: full [0, 0, 960, 623] score 170.68

• Source frame: slide-012.jpg
• Crop: full [0, 0, 960, 623] score 170.22

• Source frame: slide-013.jpg
• Crop: full [0, 0, 960, 623] score 169.26

• Source frame: slide-014.jpg
• Crop: full [0, 0, 960, 623] score 171.1

• Source frame: slide-015.jpg
• Crop: full [0, 0, 960, 623] score 174.05

Dense Scene-Detected Slide Candidates

• youtube 5_QWh4LGoxg dense slides