scholarly journals Real Time Detection and Identification of Human Emotions through Live Streaming

2019 ◽  
Vol 8 (4) ◽  
pp. 12842-12845
Keyword(s):  
Skin Color ◽  
Communication Complexity ◽  
Opinion Mining ◽  
Training Data ◽  
Live Streaming ◽  
Data Set ◽  
Detection And Identification ◽  
Challenging Tasks ◽  
The Face ◽  
Video Input

Automating the analysis of facial expressions of individuals is one of the challenging tasks in opinion mining. In this work, the proposed technique for identifying the face of an individual and the emotions, if present from a live camera. Expression detection is one of the sub-areas of computer visions which is capable of finding a person from a digital image and identify the facial expression which are the key factors of nonverbal communication. Complexity involves mainly in two cases viz., 1)if more than one emotions coexist on a face. 2) expressing same emotion between individuals is not exactly same. Our aim was to make the processes automatic by identify the expressions of people in a live video. In this system OpenCV library containing face recognizer module for detecting the face and for training the model. It was able to identify the seven different expressions with 75-85% accuracy. The expressions identified are happy, sadness, disgust, fear, anger, surprise and neutral. The this an image frame from is captured from the video, locate the face in it and then test it against the training data for predicting the emotion and update the result. This process is continued till the video input exists. On top of this the data set for training should be in such a way that , it prediction should be independent of age, gender, skin color orientation of the human face in the video and also the lamination around the subject of reference

scholarly journals ANALYSIS OF HUMAN FACE RECOGNITION ALGORITHM USING PCA+FDIT IN IMAGE DATABASE FOR CRIME INVESTIGATION

2013 ◽  
Vol 4 (3) ◽  
pp. 788-796
Author(s):  
V. S. Manjula
Keyword(s):  
Face Recognition ◽  
Real Time ◽  
Face Detection ◽  
Skin Color ◽  
Back Propagation ◽  
Real Life ◽  
Hierarchical Architecture ◽  
Detection And Identification ◽  
The Face ◽  
Real Time Application

In general, the field of face recognition has lots of research that have put interest in order to detect the face and to identify it and also to track it. Many researchers have concentrated on the face identification and detection problem by using various approaches. The proposed approach is further very useful and helpful in real time application. Thus the Face Detection, Identification  which is proposed here is used to detect the faces in videos in the real time application by using the FDIT (Face Detection Identification Technique) algorithm. Thus the proposed mechanism is very help full in identifying individual persons who are been involved in the action of robbery, murder cases and terror activities. Although in face recognition the algorithm used is of histogram equalization combined with Back propagation neural network in which we recognize an unknown test image by comparing it with the known training set images that are been stored in the database. Also the proposed approach uses skin color extraction as a parameter for face detection. A multi linear training and rectangular face feature extraction are done for training, identifying and detecting.   Thus the proposed technique   is PCA + FDIT technique configuration only improved recognition for subjects in images are included in the training data.   It is very useful in identify a single person from a group of faces.   Thus the proposed technique is well suited for all kinds faces frame work for face detection and identification. The face detection and identification modules share the same hierarchical architecture. They both consist of two layers of classifiers, a layer with a set of component classifiers and a layer with a single combination classifier.  Also we have taken a real life example and simulated the algorithms in IDL Tool successfully.

Hand gesture recognition via image processing techniques and deep CNN

2020 ◽  
Vol 39 (3) ◽  
pp. 4405-4418
Author(s):  
Yao-Liang Chung ◽  
Hung-Yuan Chung ◽  
Wei-Feng Tsai
Keyword(s):  
Skin Color ◽  
Recognition Rate ◽  
Region Of Interest ◽  
Training Data ◽  
Background Information ◽  
Hand Gestures ◽  
Data Set ◽  
Practical Applications ◽  
Deep Cnn ◽  
Hand Region

In the present study, we sought to enable instant tracking of the hand region as a region of interest (ROI) within the image range of a webcam, while also identifying specific hand gestures to facilitate the control of home appliances in smart homes or issuing of commands to human-computer interaction fields. To accomplish this objective, we first applied skin color detection and noise processing to remove unnecessary background information from the captured image, before applying background subtraction for detection of the ROI. Then, to prevent background objects or noise from influencing the ROI, we utilized the kernelized correlation filters (KCF) algorithm to implement tracking of the detected ROI. Next, the size of the ROI image was resized to 100×120 and input into a deep convolutional neural network (CNN) to enable the identification of various hand gestures. In the present study, two deep CNN architectures modified from the AlexNet CNN and VGGNet CNN, respectively, were developed by substantially reducing the number of network parameters used and appropriately adjusting internal network configuration settings. Then, the tracking and recognition process described above was continuously repeated to achieve immediate effect, with the execution of the system continuing until the hand is removed from the camera range. The results indicated excellent performance by both of the proposed deep CNN architectures. In particular, the modified version of the VGGNet CNN achieved better performance with a recognition rate of 99.90% for the utilized training data set and a recognition rate of 95.61% for the utilized test data set, which indicate the good feasibility of the system for practical applications.

scholarly journals UXO detection and identification based on intrinsic target polarizabilities — A case history

Geophysics ◽  
2009 ◽  
Vol 74 (1) ◽  
pp. B1-B8 ◽  
Author(s):  
Erika Gasperikova ◽  
J. Torquil Smith ◽  
H. Frank Morrison ◽  
Alex Becker ◽  
Karl Kappler
Keyword(s):  
Electromagnetic Induction ◽  
Current Practice ◽  
Cost Effective ◽  
Scrap Metal ◽  
Training Data ◽  
Time Dependent ◽  
Unexploded Ordnance ◽  
Data Set ◽  
Detection And Identification ◽  
Mode Of Operation

Electromagnetic induction data parameterized in time-dependent object intrinsic polarizabilities can discriminate unexploded ordnance (UXO) from false targets (scrap metal). Data from a cart-mounted system designed to discriminate UXO of [Formula: see text] in diameter are used. Discriminating UXO from irregular scrap metal is based on the principal dipole polarizabilities of a target. Nearly intact UXO displays a single major polarizability coincident with the long axis of the object and two equal, smaller transverse polarizabilities, whereas metal scraps have distinct polarizability signatures that rarely mimic those of elongated symmetric bodies. Based on a training data set of known targets, objects were identified by estimating the probability that an object is a single UXO. Our test survey took place on a military base where [Formula: see text] mortar shells and scrap metal were present. We detected and correctly discriminated all [Formula: see text] mortars, and in that process we added 7% and 17%, respectively, of dry holes (digging scrap) to the total number of excavations in two different survey modes. We also demonstrated a mode of operation that might be more cost effective than current practice.

scholarly journals Real Time Face Mask Detection using Google Cloud ML and Flutter

2022 ◽  
Vol 2161 (1) ◽  
pp. 012020
Author(s):  
Sohit Kummar ◽  
Asutosh Mohanty ◽  
Jyotsna ◽  
Sudeshna Chakraborty
Keyword(s):  
Machine Learning ◽  
Cloud Service ◽  
Face Mask ◽  
Data Set ◽  
Fine Grained ◽  
Challenging Tasks ◽  
The Face ◽  
Automated Machine Learning ◽  
Precautionary Measures ◽  
Almost All

Abstract Coronavirus (Covid-19) pandemic has impacted the whole world and has forced health emergencies internationally. The contact of this pandemic has been fallen over almost all the development sectors. A lot of precautionary measures have been taken to control the Covid-19 spread, where wearing a face mask is an essential precaution. Wearing a face mask correctly has been essential in controlling the Covid-19 transmission. Moreover, this research aims to detect the face mask with fine-grained wearing states: face with the correct mask and face without mask. Our work has two challenging tasks due to two main reasons firstly the presence of augmented data set available in the online market and the training of large datasets. This paper represents a mobile application for face mask detection. The fully automated Machine Learning Cloud service known as Google Cloud ML API is used for training the model in TensorFlow file format. This paper highlights the efficiency of the ML model. Additionally, this paper examines the advancement of the cloud technology used for machine learning over the traditional coding methods.

scholarly journals Robust Wireless Communication for Small Exploration Rovers Equipped with Multiple Antennas by Estimating Attitudes of Rovers in Several Experimental Environments

2017 ◽  
Vol 29 (5) ◽  
pp. 864-876 ◽  
Author(s):  
Masahiko Mikawa ◽  
Keyword(s):  
Estimation Method ◽  
Attitude Estimation ◽  
The Body ◽  
Training Data ◽  
Large Area ◽  
Data Set ◽  
Wireless Mesh ◽  
Surface Exploration ◽  
The Face ◽  
Micro Gravity

We are developing a robotic system for an asteroid surface exploration. The system consists ofmultiplesmall size rovers, that communicate with each other over a wireless network. Since the rovers configure over a wireless mesh sensor network on an asteroid, it is possible to explore a large area on the asteroid effectively. The rovers will be equipped with a hopping mechanism for transportation, which is suitable for exploration in a micro-gravity environment like a small asteroid’s surface. However, it is difficult to control the rover’s attitude during the landing. Therefore, a cube-shaped rover was designed. As every face has two antennas respectively, the rover has a total of twelve antennas. Furthermore, as the body shape and the antenna arrangements are symmetric, irrespective of the face on top, a reliable communication state among the rovers can be established by selecting the proper antennas on the top face. Therefore, it is important to estimate which face of the rover is on top. This paper presents an attitude estimation method based on the received signal strength indicators (RSSIs) obtained when the twelve antennas communicate among each other. Since the RSSI values change depending on an attitude of the rover and the surrounding environment, a significantly large number of RSSIs were collected as a training data set in different kinds of environments similar to an asteroid; consequently, a classifier for estimating the rover attitude was trained from the data set. A few of the experimental results establish the validity and effectiveness of the proposed exploration system and attitude estimation method.

scholarly journals Optimizing Few-Shot Learning based on Variational Autoencoders

2021 ◽  
Author(s):  
Ruoqi Wei ◽  
Ausif Mahmood
Keyword(s):  
Face Recognition ◽  
Data Augmentation ◽  
Training Data ◽  
The Novel ◽  
Limited Data ◽  
Data Set ◽  
Learning Tasks ◽  
The Face ◽  
Face Generation ◽  
Generative Approach

Despite the importance of few-shot learning, the lack of labeled training data in the real world, makes it extremely challenging for existing machine learning methods as this limited data set does not represent the data variance well. In this research, we suggest employing a generative approach using variational autoencoders (VAEs), which can be used specifically to optimize few-shot learning tasks by generating new samples with more intra-class variations. The purpose of our research is to increase the size of the training data set using various methods to improve the accuracy and robustness of the few-shot face recognition. Specifically, we employ the VAE generator to increase the size of the training data set, including the basic and the novel sets while utilizing transfer learning as the backend. Based on extensive experimental research, we analyze various data augmentation methods to observe how each method affects the accuracy of face recognition. We conclude that the face generation method we proposed can effectively improve the recognition accuracy rate to 96.47% using both the base and the novel sets.

scholarly journals Semi-Supervised Learning to Perceive Children's Affective States in a Tablet Tutor

2020 ◽  
Vol 34 (09) ◽  
pp. 13350-13357
Author(s):  
Mansi Agarwal ◽  
Jack Mostow
Keyword(s):  
Supervised Learning ◽  
Affective State ◽  
Training Data ◽  
Affective States ◽  
Data Set ◽  
Tutoring Systems ◽  
Screen Capture ◽  
Student Affect ◽  
The Face ◽  
Authentic Data

Like good human tutors, intelligent tutoring systems should detect and respond to students' affective states. However, accuracy in detecting affective states automatically has been limited by the time and expense of manually labeling training data for supervised learning. To combat this limitation, we use semi-supervised learning to train an affective state detector on a sparsely labeled, culturally novel, authentic data set in the form of screen capture videos from a Swahili literacy and numeracy tablet tutor in Tanzania that shows the face of the child using it. We achieved 88% leave-1-child-out cross-validated accuracy in distinguishing pleasant, unpleasant, and neutral affective states, compared to only 61% for the best supervised learning method we tested. This work contributes toward using automated affect detection both off-line to improve the design of intelligent tutors, and at runtime to respond to student affect based on input from a user-facing tablet camera or webcam.

scholarly journals Product Sentiment Assessment using Large Scale Cloud System

2019 ◽  
Vol 8 (12S) ◽  
pp. 1076-1080
Keyword(s):  
Feature Selection ◽  
Large Scale ◽  
Opinion Mining ◽  
Information Gain ◽  
Training Data ◽  
Support Vector ◽  
Data Set ◽  
Selection For ◽  
Support Decision Making ◽  
Original Feature

A typical manner in which valuable information can be obtained by means of extracting the sentiment or also the opinion from any message is called sentiment analysis. The sentiment classification exploits the technologies in machine learning owing to their ability to learn from training data set to predict and support decision making with high accuracy level. Some algorithms do not maintain proper scalability for large datasets. Today, there are several disciplines that have the need to deal with some big datasets for involving features in high numbers. The methods of feature selection have been aiming at the elimination of the noisy, the irrelevant or the redundant features that can bring down the performance of classification. Most of the traditional methods lack the scalability to be able to cope with the results within a given time. Here in this work, Term Frequency (TF) that is a method of feature extraction has been used. The focus has been on the selection for the opinion mining by using the Information Gain (IG) based method and compared with the method of. All these methods of feature selection have reduced all the original feature sets by means of removing the features that are irrelevant to enhance the accuracy of classification and bring down the running time of the learning algorithms. The method proposed has been evaluated by means of using the Support Vector Machine (SVM) based classifier. The experimental results have proved that the proposed method had achieved better performance.

A driver drowsiness and distraction warning system based on raspberry Pi 3 Kit

2019 ◽  
Vol 70 (3) ◽  
pp. 184-192
Author(s):  
Toan Dao Thanh ◽  
Vo Thien Linh
Keyword(s):  
Processing System ◽  
Warning System ◽  
Raspberry Pi ◽  
Data Set ◽  
Lower Lip ◽  
Driver Drowsiness ◽  
Image Sensing ◽  
The Face ◽  
The Common ◽  
University Laboratory

In this article, a system to detect driver drowsiness and distraction based on image sensing technique is created. With a camera used to observe the face of driver, the image processing system embedded in the Raspberry Pi 3 Kit will generate a warning sound when the driver shows drowsiness based on the eye-closed state or a yawn. To detect the closed eye state, we use the ratio of the distance between the eyelids and the ratio of the distance between the upper lip and the lower lip when yawning. A trained data set to extract 68 facial features and “frontal face detectors” in Dlib are utilized to determine the eyes and mouth positions needed to carry out identification. Experimental data from the tests of the system on Vietnamese volunteers in our University laboratory show that the system can detect at realtime the common driver states of “Normal”, “Close eyes”, “Yawn” or “Distraction”

Cloning Safe Driving Behavior for Self-Driving Cars using Convolutional Neural Networks

2019 ◽  
Vol 12 (2) ◽  
pp. 120-127 ◽  
Author(s):  
Wael Farag
Keyword(s):  
Gradient Descent ◽  
Autonomous Driving ◽  
Driving Behavior ◽  
Training Data ◽  
Stochastic Gradient Descent ◽  
Data Set ◽  
Safe Driving ◽  
Processing Pipeline ◽  
Self Driving Cars ◽  
And Training

Background: In this paper, a Convolutional Neural Network (CNN) to learn safe driving behavior and smooth steering manoeuvring, is proposed as an empowerment of autonomous driving technologies. The training data is collected from a front-facing camera and the steering commands issued by an experienced driver driving in traffic as well as urban roads. Methods: This data is then used to train the proposed CNN to facilitate what it is called “Behavioral Cloning”. The proposed Behavior Cloning CNN is named as “BCNet”, and its deep seventeen-layer architecture has been selected after extensive trials. The BCNet got trained using Adam’s optimization algorithm as a variant of the Stochastic Gradient Descent (SGD) technique. Results: The paper goes through the development and training process in details and shows the image processing pipeline harnessed in the development. Conclusion: The proposed approach proved successful in cloning the driving behavior embedded in the training data set after extensive simulations.

Sign in / Sign up

Export Citation Format

Share Document