Tech Talk on Artificial Intelligence by Navneet@Snapshopr

Artificial Intelligence By Navneet@Snapshopr

Introduction

Artificial Intelligence(AI) is the field of study in which we try to build intelligent systems that can make decisions, identify things and understand natural language similar to how humans do.

Subfields of AI

● Machine Learning

● Computer Vision

● Natural Language Processing

Machine Learning

● The field of AI that helps computers act without being explicitely programmed

● Design and study of systems that can learn from data.

Computer Vision

● The field of AI that tries to mimic the vision system of the brain using learning algorithms.

● Face Recognition,Object Recognition and Scene Understanding are some of the popular problems in this domain.

Natural Language Processing

● The field of AI that deals with understanding natural language.

● Text Processing, Language Translation, Speech Recognition are some of the popular problems in this domain.

Computer Vision : My Focus@Snapshopr

● “If we want machines to think, we need to teach them to see.”

- Fei-Fei Li

Modern AI : Deep Learning

Advanced machine learning that requires huge amount of data and uses giant neural networks.

People Behind It :LeCun, Hinton, Bengio & Ng

Closer Look at Deep Learning

State of the Art

● Some of the most challenging problems in computer vision were solved with promising results in 2014.

● They all made use of Deep Learning as part of the overall system design.

Breakthrough One

Captioned by Human and by Stanford & Google’s Experimental Program

Human: “A person riding a dirt bike is covered in mud.” Computer model: “A person riding a motorcycle on a dirt road.”

Breakthrough Two

● DeepFace by Facebook AI Research

DeepFace uses a 3-D model to rotate faces, virtually, so that they facethe camera. Image (a) shows the original image, and(g) shows the final, corrected version.

Breakthrough Three

Yahoo Labs in California and Stanford University revealed a new approach to the problem of spotting faces at an angle - including upside down - even when partially occluded.

Research at Snapshopr

● Snapshopr is a cutting-edge deep learning startup in Bangalore.Its first product is trying to solve fashion search on mobile.

● Apart from our market facing product, we invest our time and resources as part of our R&D unit Snapshopr X.

Snapshopr X : Pattern Recognition

Snapshopr X : Object Recognition

Risks of AI

● "The development of full artificial intelligence could spell the end of the human race." - Stephen Hawking

Risks of AI

● “In the longer term, AI could be our biggest existential threat." - Elon Musk

Disruptive Automation

● Automation across industries powered by technologies such as AI is likely to cause societal issues as many routine jobs are well on way to be replaced by software.

● It's advised by experts that people should be prepared for this impending change and be ready to build skills to remain employable.

Path Ahead ...

The immediate usefullness of AI far outweighs its distant risks and hence a controlled approach towards the development of AI is more advisable and can help us in fields as diverse as medicine, energy, automobile and retail, and contribute positively to the society.

Thank You

@NavSha

@SnapShopr