An Optimized Deep Vision Framework

(DETECTRON)

Authors

  • Ankur Chaturvedi, Dr. Vikram Rajpoot

Abstract

DEEP VISION (DETECTRON) is a project that is intended to be useful for many other computer vision research institutions. It takes desired neural schema and training data as input and creates high quality deep learning models for computer vision requirements. User doesn’t necessarily need a lot of knowledge of machine learning, rather he can just annotate his input images for training and give his desired neural schema and DETECTRON takes care of rest. Apart from creating desired Deep Learning models, it can provide shorter and faster neural networks using our reliable optimization algorithm. Mainly the whole project/system is divided into 5 major modules:

Published

2020-06-30

Issue

Section

Articles