Scribd
Upload
3K views

PPT

This document outlines an agenda for a deep learning project including an introduction to deep learning, a literature survey, problem statement, objectives, proposed methodology, and references. The proposed methodology section describes using a recurrent neural network for input data preprocessing, detection of objects in images/audio/text, and subjective and objective analysis of results. MATLAB deep learning and neural network toolboxes would be used, with an example of deep learning included. [/SUMMARY]

Uploaded by

VITHAL BABLE
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
3K views

PPT

This document outlines an agenda for a deep learning project including an introduction to deep learning, a literature survey, problem statement, objectives, proposed methodology, and references. The proposed methodology section describes using a recurrent neural network for input data preprocessing, detection of objects in images/audio/text, and subjective and objective analysis of results. MATLAB deep learning and neural network toolboxes would be used, with an example of deep learning included. [/SUMMARY]

Uploaded by

VITHAL BABLE
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 21

AGENDA

 Introduction
 Literature Survey
 Problem Statement
 Objectives
 Proposed Methodology
 References
Introduction
 Deep learning is the hottest field in AI right now.
 From Google Duplex assistant to Tesla self- driving cars the
applications are endless.
 In the past 10 years, machine learning and artificial intelligence
have shown tremendous progress.
 Recently deep learning plays major role for
1.Explosion of huge data
2.Cheap Computing cost
3.Improvement in ML algorithms
INTRODUCTION
 Deep learning allows computational models that are composed
of multiple processing layers to learn representations of data
with multiple levels of abstraction.
 These methods have dramatically improved the state-of-the-art
in speech recognition, visual object recognition, object detection
and many other domains such as drug discovery and genomics.
 Deep learning discovers intricate structure in large data sets by
using the backpropagation algorithm to indicate how a machine
should change its internal parameters that are used to compute
the representation in each layer from the representation in the
previous layer.
 Deep convolutional nets have brought about breakthroughs in
processing images, video, speech and audio, whereas recurrent
nets have shone light on sequential data such as text and sp
 Conventional machine-learning techniques were limited in
their ability to process natural data in their raw form.
Machine learning and Deep Learning Difference
PROPOSED METHOD

Input data

Preprocess

RNN

Detected Object
Proposed
Recurrent Neural Networks
Contd…

1.Input Image/Audio/Text data


2.Preprocessing(Resizing/converting in double)
3.Apply Deep Neural Network
4. Detected Object
5.Subjective and objective analysis.
Technology Used : MATLAB

MATLAB (matrix laboratory) is a multi-


paradigm numerical computing environment
and fourth-generation programming
language.
Its one of the easiest and most productive
software development environment for
engineers.
MATLAB toolboxes Used

 Deep Learning Toolboxes


 Convolution Neural
Networks
 Autoencoders for deep
learning toolboxes.

Example of DEEP LEARNING


ADVANTAGES

1. Time complexity is less


2. Robust : natural variations in data is automatically
learned.
3. Generalizable : Same DNN can be used for different
applications.
4. Scalability: If we increase the data then also we
will get the same algorithm will work on that data
without any further modifications in algorithm.
Basic Weakness

 Deep learning require large dataset hence it takes


much time for training.
 Some features are not human understandable.
 Requires good understanding for modelling.
Experimental Analysis

 There are two ways to analyze the Deep learning results


and compared with state of art techniques,
1. Subjective Analysis
2. Objective Analysis
Subjective Analysis

 By observing the results by HVS we can conclude the


experiment improvement using deep learning.
 Subjective analysis may vary person to person so we are
using objective analysis also.
CONCLUSION
 In this article we introduced deep learning, a subfield of
machine learning with applications in many industries as
well as one of the hottest topics in research nowadays.
Applications
References

1. Krizhevsky, A., Sutskever, I. & Hinton, G. ImageNet classification with


deep convolutional neural networks. In Proc. Advances in Neural
Information Processing Systems 25 1090–1098 (2012).
2. Farabet, C., Couprie, C., Najman, L. & LeCun, Y. Learning hierarchical
features for scene labeling. IEEE Trans. Pattern Anal. Mach. Intell. 35,
1915–1929 (2013).
3. Tompson, J., Jain, A., LeCun, Y. & Bregler, C. Joint training of a
convolutional network and a graphical model for human pose
estimation. In Proc.
4. Szegedy, C. et al. Going deeper with convolutions. Preprint at
http://arxiv.org/ abs/1409.4842 (2014).
5. Mikolov, T., Deoras, A., Povey, D., Burget, L. & Cernocky, J. Strategies for
training large scale neural network language models.

You might also like