FACIAL RECOGNITION

USING MACHINE LEARNING

ABSTRACT
 Automatic recognition of human affects has become more interesting and challenging
problem in artificial intelligence, human-computer interaction and computer vision
fields.
 Facial Recognition (FE) is the one of the most significant features to recognize the
emotion of human in daily human interaction.
 FE Recognition (FER) has received important interest from psychologists and computer
scientists for the applications of health care assessment, human affect analysis, and
human computer interaction.
 Human express their emotions in a number of ways including body gesture, word, vocal
and facial Recognitions.
 Recognition is the important channel to convey emotion information of different people
because face can express mainly human emotion.
 This project surveys the current research works related to facial Recognition recognition.
 The study attends to explored details of the facial datasets, feature extraction methods,
the comparison results and futures studies of the facial emotion system.
INTRODUCTION
 For facial Recognition recognition problems, the general recognition method
appeared in previous work can be divided into two major steps, face
representation and classifier construction.
 In the first step, features related to facial Recognition are extracted from
images.
 Some of the features are hand-designed, whereas others are learnt from
training images.
 Then, the dimensionality of the features is reduced to facilitate an efficient
classification and enhance the generalization capability.
 The universal Recognitions which are mentioned in the papers are usually
anger, disgust, fear, joy (or happiness), sadness and surprise, whereas some
researchers add neutral as the seventh Recognition.
 In the second step, classifiers are designed based on the simplified features to
classify each Recognition as one of the six (or seven) Recognitions.
LITERATURE SURVEY 1
 Implementation of Facial Recognition Recognition System for
Selecting Fashion Item Based on Like and Dislike Recognition
Lutfiyatul Fatjriyati Anas ; IEEE 2021.
 This study was developed to find out what opinion or assessment of a
person to a picture, especially fashion images through the Recognition
of a person's face taken when looking at a fashion picture.
 Facial Recognitions that are processed will produce Recognitions of
like or dislike.
 This facial express recognition system is built using the Facial
Landmark to get the Recognition feature.
 Then the value of the feature is normalized using min-max.
 Finally, facial Recognitions are classified with k-Nearest Neighbor. The
results of this system show which fashion images are liked by someone.
LITERATURE SURVEY 2
 A New Method for Facial and Corporal Recognition
Recognition Kahina Amara ; IEEE 2021.
 This work concentrates on facial and corporal

Recognition and feature extraction.

 For system performance assessment, we used leave-

one-out subject cross-validation.

 The obtained results show the superior performance of

the RGB-D features provided by Kinect (v2).

LITERATURE SURVEY 3
 Facial Recognition Recognition by Calculating Euclidian Distance
for Eigen Faces Using PCA Mangala B.S Divya ; N.B Prajwala
IEEE 2021.
 In this paper the proposed methodology is to recognize Facial
Recognitions of human being by using Eigen Faces.
 For recognition of human Recognition, in this approach we consider
the following steps: face detection, facial feature extraction and then
facial Recognition classification.
 The proposed system is based on calculating the Euclidian distance
for the Eigen Faces.
 In this approach we will consider all the seven basic emotions which
are classified as happy, anger, sad, fear, and disgust, surprised, neutral.
LITERATURE SURVEY 4
 Effects of Facial Mood Recognitions on Face Biometric
Recognition System’s Reliability Mohammadreza Azimi IEEE
2021.
 In the study, the local binary pattern (LBP) attributes obtained from
the surrounding of the landmarks using active shape models are used.
 In order to find the most discriminating subset of the obtained
attributes, the selection of the attributes has been applied for improve
the recognition rate.
 It has been shown that the method proposed in experiments with 10-
fold cross-validation with the Cohn-Kanade dataset (CK+) which is
containing seven different Recognition classes achieves %89.71
success rate.
LITERATURE SURVEY 5
 Recognizing Fine Facial Micro-Recognitions Using Two-
Dimensional Landmark Feature Dong Yoon Choi ; IEEE 2021.
 In this paper, we propose a two-dimensional (2D) landmark feature
for effectively recognizing facial micro-Recognition.
 The proposed 2D landmark feature is obtained by converting existing
coordinate-based landmark information into 2D image information,
and has an advantage of having a unique feature according to
emotions regardless of the intensity of facial Recognition.
 Thus, we can achieve effective emotion recognition by learning the
proposed 2D landmark feature information on a convolutional neural
network (CNN) and a long-term term memory (LSTM)-based
network.
EXISTING SYSTEM
 In Existing System Edge Detection Algorithm is used
which actually segments only the edges of the rice that
we are going to analyze.
 In existing system for identifying emotion in face they
have used Edge Detection Technique.
 The main drawback of this system is only the edges of
faces are been segmented because of this inner areas
which actually shows the emotion cannot be segmented.
 Over lapping of images are also high which will increase
the SNR ratio which inturn reduces the accuracy.
EXISTING SYSTEM DISADVANTAGES
 Prediction of image size is not accurate.
 Time consumption.
 Noise will be high.
 Overlapping of images will be high.
PROPOSED SYSTEM
 We propose a method to search optimized active regions from the active
regions in face.
 A CNN is trained for each kind of optimized active regions to extract
features and classify Recognitions.
 In order to get representable features, histogram equalization, rotation
correction and spatial normalization are carried out on the Recognition
images.
 A decision-level fusion method is applied, by which the final result of
Recognition recognition is obtained via majority voting of the three results.
 Experiments on both independent databases and fused database are carried
out to evaluate the performance of the proposed system.
 Our novel method achieves higher accuracy compared to previous
literature, with the added benefit of low latency for inference.
PROPOSED SYSTEM ADVANTAGES
 Prediction of image size is accurate.
 Time consumption is less.
 Noise will be less.
 No Overlapping of images.
ARCHITECTURE DIAGRAM
HARDWARE REQUIREMENTS

 Processor : Intel i3 or later

 Hard Disk : 500 GB
 RAM : 4 GB
 Operating System : Windows7 or later
SOFTWAREWARE REQUIREMENTS

 Technology : Python
 IDE : Python IDE
 Database : My SQL
MODULES
 Image Acquisition
 Image Preprocessing
 Image Segmentation
 Feature Extraction
 Classification.
MODULE DESCRIPTION
 1. Image Acquisition Image
 Acquisition is the process of collection of images. We had
performed a refined search on the dataset of facial Recognitions
but could not find one that matches our requirements.
 2. Image Preprocessing
 Image preprocessing includes converting RBG images into
Grayscale images using Python.
 An RGB image means the images present with its original colors.
Grayscale images have the combination of black and white.
 Conversion of RGB to grayscale is done for enhancing the dataset
available. Converting the images to grayscale helps in improving
the accuracy of the result.
MODULE DESCRIPTION
 3. Image Segmentation
 Image segmentation breaks the image down into meaningful regions. It divides
digital image into multiple segments.
 The goal is to simplify or change the representation into more meaningful image. It
differentiates between the objects we want to inspect further and the other objects
or their background.
 4. Feature Extraction
 Feature extraction is extracting or showing of the face portion of the head so that
classification becomes easy. Features are extracted in order to differentiate
between the images.
 Features extraction is used in almost all machine vision algorithms. The common
goal of feature extraction and representation techniques is to convert the
segmented objects into representations that better describe their main features and
attributes.
 Here, shape of the spots or the eyebrows and lip area of the face is extracted.
MODULE DESCRIPTION
 5. Classification
 The last module includes the classification in which Tensor Flow
and Machine Learning algorithm will be used.
 Tensor Flow is a Python-friendly open source library for
numerical computation that makes machine learning faster and
easier.
 Tensor Flow allows developers to create dataflow graphs -
structures that describe how data moves through a graph, or a
series of processing nodes.
 Each node in the graph represents a mathematical operation, and
each connection or edge between nodes is a multidimensional
data array, or tensor.
CONCLUSION
 Numerous researches and studies about Emotion Recognition, Machine
learning and Deep learning techniques used for recognizing the
emotions are conducted.
 It is required in future to have a model like this with much more
reliable, which has limitless possibilities in all fields.
 This project tried to use inception net for solving emotion recognition
problem.
 Various databases have been explored, Kaggle’s and Karolinska
Directed Emotional Faces (KDEF) is used as dataset for carrying out
the research.
 Tensor Flow is used to train the model.
 Good Accuracy rate is achieved. In future, real time emotion
recognition can be developed using the same architecture.
REFERNCES
 1. Farahani, Fatemeh Shahrabi, Mansour Sheikhan, and Ali Farrokhi. "A fuzzy approach
for facial emotion recognition." 2013 13th Iranian Conference on Fuzzy Systems
(IFSC). IEEE, 2013
 2. Oh, Byung-Hun, and Kwang-Seok Hong. "A study on facial components detection
method for face-based emotion recognition." 2014 International Conference on Audio,
Language and Image Processing. IEEE, 2014
 3. Reney, Dolly, and Neeta Tripathi. "An Efficient Method to Face and Emotion
Detection." 2015 Fifth International Conference on Communication Systems and
Network Technologies. IEEE, 2015.
 4. N. Cristiana, T. Shawe, An Introduction to Support Vector Machine, Cambridge
University Press, 2000.
 5. Pantic, Maja, and Leon JM Rothkrantz. "Toward an affect-sensitive multimodal
human-computer interaction" Proceedings of the IEEE 91.9 (2003): 1370-1390.
 6. Adeyanju, Ibrahim A., Elijah O. Omidiora, and Omobolaji F. Oyedokun.
"Performance evaluation of different support vector machine kernels for face emotion
recognition." 2015 SAI Intelligent Systems Conference (IntelliSys). IEEE, 2015.