Page 58 - Flytxt

ADVANCING MACHINE VISION:

THE CHALLENGES

With the advent of Artificial Intelligence, machines can potentially
work exactly like the human mind. While deep learning has
emerged as a powerful tool to perfect machine perception,

machines are yet to achieve accuracy in identifying different

objects and environments. Flytxt R&D team is discussing their
approach to creating an end-to-end analytics pipeline for scene
recognition. Here’s more.

- Jobin Wilson, Principal Data Scientist - R&D and
Muhammad Arif, Lead Engineer, Flytxt

he human brain can carry out Image classification is one of the

some amazing tasks such as hallmark tasks of computer vision. It

Tunderstanding the world in a allows defining a context for object
single visual frame. It takes only a few recognition. This will have diverse

tens of milliseconds for the brain to applications. However, the classic

recognize an object or environment. problem in computer vision is that
of determining whether image data
Furthermore, humans are capable of

learning and remembering a diverse
set of places and patterns, and solving

complex problems such as planning and The Data Science R&D team

navigation, involving vision, perception at Flytxt has released an end-
and cognition. The human neural to-end scene recognition

architecture has inspired researchers pipeline consisting of feature
to simulate such abilities on machines extraction, encoding, pooling

to solve challenging problems using and classification. The primary

artificial intelligence. Through artificial objective of this project
intelligence, machines have come closer is to clearly outline the

to human ability in several cognitive tasks practical implementation

such as visualizing and identifying diverse of a basic scene-recognition
objects and environments. Consequently, pipeline having reasonable

deep learning has emerged as a powerful accuracy, using conventional

tool to solve problems involving machine computer vision techniques.
vision and perception.

58 INSIGHTZ - VOLUME 03, 2018