Scribd
Upload
204 views

Emotion Detection

The document discusses emotion recognition using deep neural networks and convolutional neural networks. It describes how CNNs can be used for feature extraction from facial images through techniques like convolution and max pooling. The methodology involves preprocessing input images, extracting features from facial action units, and classifying emotions. Challenges in emotion recognition from images include pose, head movements, occlusions, and ambiguity. Potential applications are discussed like using emotion recognition in cars, health, education, market research, and interviews. The conclusion discusses future scope in these domains. References include research papers and external resources on CNNs, emotion recognition, and datasets.

Uploaded by

het patel
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
204 views

Emotion Detection

The document discusses emotion recognition using deep neural networks and convolutional neural networks. It describes how CNNs can be used for feature extraction from facial images through techniques like convolution and max pooling. The methodology involves preprocessing input images, extracting features from facial action units, and classifying emotions. Challenges in emotion recognition from images include pose, head movements, occlusions, and ambiguity. Potential applications are discussed like using emotion recognition in cars, health, education, market research, and interviews. The conclusion discusses future scope in these domains. References include research papers and external resources on CNNs, emotion recognition, and datasets.

Uploaded by

het patel
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 17

Emotion Recognition

System
Content
Part 01 Introduction

Part 02 Requirement Specification

Part 03 Conclusion and Future Scope

Part 04 References
Introduction
• As Human-Robot interaction is increasing its attention nowadays.
• In order to put some limelight on socializing robots with humans,
understanding the facial gestures and visual cues of an individual is a
need.
It allows a robot to understand the expressionof human in turn enhancing its
effectiveness in performing various tasks.
Socially intelligent software tools can be accomplished.
It serves as a mearsurement system for behavioural tools.
Honest Feedbacks Resources and
Tools

1 6

ML/AI Trend
2 Sodhyatra Counselling
5

3 4
Entertainment Depression
DNN(Deep Neural Network)
• A neural network, in general, is a technology built to simulate the activity of the
human brain – specifically, pattern recognition and the passage of input through
various layers of simulated neural connections.
• Many experts define deep neural networks as networks that have an input layer,
an output layer and at least one hidden layer in between. Each layer performs
specific types of sorting and ordering in a process that some refer to as “feature
hierarchy.” One of the key uses of these sophisticated neural networks is dealing
with unlabeled or unstructured data. The phrase “deep learning” is also used to
describe these deep neural networks, as deep learning represents a specific form
of machine learning where technologies using aspects of artificial intelligence
seek to classify and order information in ways that go beyond simple
input/output protocols.
Emotion Recognition using CNN(Convolution
Neural Network)
• Convolution Neural Network or CNN as it is popularly called is a
collection of two types of layers-
• The hidden layers / Feature Extraction Part
• Convolutions
• Pooling
• The classifier part
• Convolution is a mathematical operation which involves a combination of two
functions to produce a third function. In CNN the convolution is performed on the
input data with the use of a filter to produce a feature map.
• Pooling
Pooling layer is added after a convolution layer. It performs continuous
dimensionality reduction i.e. reduces the number of parameters and
computations thereby shortening training time and controlling overfitting. One
such pooling technique is called max-pooling, which takes the maximum value
in each window which decreases the feature map size while keeping the
significant information.
Methodology

Input storage data


1. Static Image
2. Sequence Video
Pre-processing
1. Head pose Identification
2. Face Tracking
3. Facial part recognition
Feature Extraction
1. Localization of facial action units
2. Facial point tracking
3. Evolving feature points Feature Classification
1. Using Machine Learning
2. Statistical Methods

Emotion Recognition
The Emotions can be classified into many type
and can be determined through facial
expressions.

Happy Calm Disgusted


Sad Unknown Angry
Confused Fear Surprised
Requirement Specification
• Hardware Requirement
• A powerful GPU/ CPU in the likes of Nvidia GeForce RTX 2080 Ti/ GTX 1060Ti, is required
to train the model.

• Software Requirement
• Python language and its libraries(Numpy, Pandas, Keras, etc) are required along with
latest IDE. To train the model, instead of our own resources, we can use cloud services
like GCP(Google Cloud Platform), AWS, Jupyter Notebooks, Google Collab, etc.
Challenges faced in Recognising
Emotions

Pose and frequent Image Orietation


head movements and Imaging
Conditions

Presence of
Occlusion Challenges structural
components

Ambiguity and Subtle facial


uncertainity in face deformation
motion measurement
Conclusion and Future Scope

Car System
Car Speed can be suggested or reduced by detecting drivers mood

Health
Psychiatrists to detect patients emotion for better treatment

Schools and colleges


For counselling of students

Market Research Companies


For getting honest feedbacks

Company Interviewer
For questioning the applicant by detecting mood.
References
Research Papers
Sr. No. Name Publisher Date Method Dataset
1. Automatic Emotion Monika Feb-2016 DNN(Deep MUG facial
Recognition Using Dubey Neural expression
Facial Expression: A Networks), database,
Review Prof. Lokesh SVM(Support PIE face
Singh vector image
machine), database,
LDA learning JAFFE and
algorithm and MIT+CMU
MspLBP database
features
2. A Brief Review of Byoung Jan-2018 convolutional JAFFE and
Facial Emotion Chul Ko neural MIT+CMU
Recognition Based networks, database,
on Visual facial action BU-3DFE,
Information coding MMI, Nova
system, Emotions
facial action
unit,
SVM
Books/Courses/Articles

Sr. No. Name Platform Author/Instructor Topics


1. Python Crash Self Eric Matthes Python Crash
Course: A Hands- Course for
On, Project-Based Beginners
Introduction to
Programming

2. Understanding of Medium Prabhu Convolution


Convolutional neural
Neural Network networks
(CNN)

3. Introduction to edx Microsoft Python


python fundamentals
4. Neural networks Coursera Andrew Ng Concepts of
and deep learning deep learning,
Neural
networks
Datasets

Sr. No. Name Description


1. Extended Cohn- 123 subject, 600 images.
Kanade Datasbase
2. MMI Facial expression 75 subject, 2900 videos (extract frames from these videos to get
detection single images)
3. Belfast Database There are 3 sets in this database. Set 3 has HD resolution
4. Face Detection in A simple, yet useful dataset, Face Detection in Images contains just
Images with Bounding over 500 images with approximately 1,100 faces already tagged
Boxes with bounding boxes.

5. UMDFaces The UMDFaces dataset has over 367,000 face annotations across
over 8,200 different subjects in still images. Apart from those
images, the dataset also includes over 3.7 million video frames all
annotated with facial keypoints of over 3,100 subjects. It should
be noted that this dataset is strictly for non-commercial research
purposes only.
THANK YOU

You might also like