How Neural Networks Find Polygons to Understand Documents Better
Data Science Conference via YouTube
50% OFF: In-Depth AI & Machine Learning Course
Build the Finance Skills That Lead to Promotions — Not Just Certificates
Overview
Google, IBM & Meta Certificates — All 10,000+ Courses at 40% Off
One annual plan covers every course and certificate on Coursera. 40% off for a limited time.
Get Full Access
Explore a 31-minute conference talk from Data Science Conference Europe 2023 that delves into document geometry analysis through neural networks. Learn about various deep learning approaches for understanding document layouts, from basic template matching to sophisticated techniques like differential rasterization and sampling-argmax. Gain valuable insights applicable to computer vision tasks involving man-made objects, including document analysis, aerial imagery, satellite data, and indoor scene understanding. Delivered by Boris Zimka in Belgrade, this presentation illuminates crucial steps in document analysis pipelines, examining the strengths and limitations of different methodologies for geometric understanding of document images.
Syllabus
How our neural networks find polygons to understand your documents better|Boris Zimka |DSC Europe 23
Taught by
Data Science Conference