Pattern recognition in computer vision using discrete structures and deep learning

Rashmi Ashtagi

doi:10.47974/JDMSC-2511

Pattern recognition in computer vision using discrete structures and deep learning

2026-02-01

0

PRE

AI

DOI：10.47974/JDMSC-2511

Original

Original request for help

Save

Abstract

En 中文

This work looks into ways to use deep learning and discrete mathematical structures together to make computer vision pattern recognition better. The suggested mixed models combine vector metric spaces, orthogonal sets, and discrete representations to make them better at finding features, being stable, and being easy to grasp. The technology uses novel data preparation tools and cutting-edge deep neural architectures that are based on discrete math's to record intricate visual patterns. An in-depth architectural design and processing flow are shown, along with custom loss functions and optimization techniques that are used in training methods. It was tested and found that this way works better and needs fewer computers power than others.

Keywords:

Pattern recognition

Discrete structures

Deep learning

Computer vision

Vector metric spaces

Pattern recognition in computer vision using discrete structures and deep learning

Abstract

Journal

Organization

Cited Papers