Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userFace Expression Classification Using Neuro-Fuzzy Controller
تصنيف تعابير الوجه باستخدام متحكم ضبابي عصبوني
2194
0 74
0
(
0
)
Created by
Shamra Editor
Ask ChatGPT about the research
الغاية من هذا المقال إلقاء الضوء على آلية و مراحل تصميم متحكم ضبابي عصبوني, يقوم بتحديد انتماء وجه مدخل إلى أي من تعابير الوجه الأربعة التالية و هي الفرح, الحزن, الغضب, و الخوف, و ذلك وفقا للنقاط المميزة في الوجه FCP المأخوذة من نصف الوجه, و المتعلقة بالعناصر الثلاث العين و الحاجب و نصف الفم, خلافا للدراسات المعهودة في هذا المجال الذي تعتمد على الوجه بالكامل.
The purpose of this article is to shed light on the mechanism and the procedures of a neuro-fuzzy controller that classifies an input face into any of the four facial expressions, which are Happiness, Sadness, Anger and Fear. This program works according to the facial characteristic points-FCP which is taken from one side of the face, and depends, in contrast with some traditional studies which rely on the whole face, on three components: Eyebrows, Eyes and Mouth.
Artificial intelligence review:
Research summary
يهدف هذا البحث إلى تسليط الضوء على آلية ومراحل تصميم متحكم ضبابي عصبوني يقوم بتصنيف تعابير الوجه إلى أربعة تعابير رئيسية: الفرح، الحزن، الغضب، والخوف. يعتمد النظام على نقاط مميزة في نصف الوجه تشمل العين، الحاجب، ونصف الفم، بدلاً من تحليل الوجه بالكامل كما هو شائع في الدراسات التقليدية. تم تدريب الشبكة العصبونية باستخدام خوارزمية الانتشار الخلفي، وأظهرت النتائج التجريبية دقة تصل إلى 90% في تمييز التعابير الأربعة الأساسية. تم استخدام صور لأشخاص من أعمار وأعراق مختلفة لتدريب النظام واختباره. يتألف النظام من عدة مراحل تشمل معالجة الصورة، استخلاص الصفات، تحديد المجموعات الضبابية، بناء المولد الضبابي، تحديد قواعد الاستدلال، وبناء المخمد الضبابي. تم اختيار طريقة Sugeno لتصميم المتحكم الضبابي العصبوني، وأظهرت النتائج أن النظام قادر على تحقيق دقة جيدة في تصنيف تعابير الوجه بناءً على تحليل نصف الوجه فقط.
Critical review
دراسة نقدية: يعد هذا البحث خطوة مهمة في مجال تمييز تعابير الوجه باستخدام تقنيات الذكاء الاصطناعي، إلا أن هناك بعض النقاط التي يمكن تحسينها. أولاً، الاعتماد على نصف الوجه فقط قد يؤدي إلى فقدان بعض المعلومات الهامة التي يمكن أن تكون موجودة في النصف الآخر من الوجه. ثانياً، على الرغم من أن دقة النظام تصل إلى 90%، إلا أن هناك مجالاً لتحسين هذه النسبة من خلال استخدام تقنيات تعلم أعمق أو دمج المزيد من البيانات المتنوعة. ثالثاً، لم يتم التطرق إلى كيفية تعامل النظام مع تعابير الوجه المعقدة أو المختلطة التي قد تحتوي على أكثر من تعبير واحد في نفس الوقت. أخيراً، يمكن تحسين الدراسة من خلال إجراء اختبارات على مجموعة أكبر من البيانات للتحقق من فعالية النظام في بيئات مختلفة وظروف إضاءة متنوعة.
Questions related to the research
-
ما هي التعابير الأربعة التي يقوم النظام بتصنيفها؟
النظام يقوم بتصنيف تعابير الوجه إلى أربعة تعابير رئيسية: الفرح، الحزن، الغضب، والخوف.
-
ما هي النقاط المميزة التي يعتمد عليها النظام في تصنيف تعابير الوجه؟
يعتمد النظام على نقاط مميزة في نصف الوجه تشمل العين، الحاجب، ونصف الفم.
-
ما هي دقة النظام في تمييز تعابير الوجه الأربعة الأساسية؟
أظهرت النتائج التجريبية أن دقة النظام تصل إلى 90% في تمييز التعابير الأربعة الأساسية.
-
ما هي الطريقة المستخدمة في تصميم المتحكم الضبابي العصبوني؟
تم استخدام طريقة Sugeno لتصميم المتحكم الضبابي العصبوني.
References used
RASOULZADEH M, 2012- Facial Expression recognition using Fuzzy Inference System. International Journal of Engineering and Innovative Technology (IJEIT) Volume 1, Issue 4, April 2012
EKMAN P, FRIESEN W, 1971- Constants across cultures in the face and emotion. Personality Social Psychol. 17 (2) pp.124– 129
EKMAN P, FRIESEN W, 1978- Facial Action Coding System: A Technique for the Measurement of Facial Movement. Consulting Psychologists Press, Palo Alto
rate research
Read More
The purpose of this article is to shed light on the mechanism
and the procedures of a program that classifies an input face into
any of the six basic facial expressions, which are Anger, Disgust,
Fear, Happiness, Sadness and Surprise, in addition
to normal face.
This program works by apply PCA- principal component
analysis algorithm, which is applied of one side of the face, and
depends, on contrast to the traditional studies which rely on the
whole face, on three components: Eyebrows, Eyes and Mouth.
Those out-value are used to determine the facial feature array as
an input to the neural network, and the neural network is trained by
using the back-propagation algorithm. Note that the faces used in
this study belong to people from different ages and races.
An ANFIS controller also designed and a comparison
between proposed controller, ANFIS controller and open loop model
had made with different types of disturbance.
This paper presents the proposed Method for designing fuzzy
supervisory controller model for Proportional Integral Differential
controller (PID) by Fuzzy Reasoning Petri Net (FRPN),the Features
of Method shows the fuzzification value for each prop
erty of
membership function for each input of fuzzy supervisory controller,
and determine the total number of rules required in designing the
controller before enter the appropriate rules in the design phase of
the rules, and determine the value of the inputs of the rule that has
been activated, and assembly variables that have the same property
and show the value for each of them programmatically, and
determine the deffuzification value using deffuzification methods.
The nonlinear model of Unmanned Aerial Vehicle( UAV) has been
recognized. Airosim Matlab toolbox has been used to guarantee a
simulation model for the Aerosonde.In the first stage, a
linearization technique is used to calculate the mathematical
m
odel of the UAV at a specific operation point, then PID controller
is used to stabilize this linear model. At the final stage, an
augmented feedback neural network adaptive controller is
applied to stabilize the overall nonlinear system.
As we enter the age of artificial intelligence, the need for intelligent home appliances has become very important for what this smart equipment can provide in the provision of electrical energy and water resources that are treasures should human pre
servation, in addition to the contribution of this equipment to protect the environment from pollution, where we face the challenges next:
High prices of electrical equipment.
The number of hours of electricity supply in many areas is low
because of the current conditions in our country.
- Water shortage.
- The rise in prices of materials used in daily life in general and
household detergents in particular
- Great waste of electricity.
- Pollution of the environment and groundwater with detergents used in the laundry process.
Moreover, the unjust economic blockade imposed on our country is pushing us to work to produce low-cost national housing equipment that competes with foreign products in order to alleviate the material burden on the citizens and promote the national economy.
In order to accomplish this smart washing machine, we have written a code for f type-2 fuzzy microcontroller, using the Python programming language. This controller has received four entries, which are: The first income (clothing color), obtained by taking a picture of the clothes that we need to wash by a camera with a resolution of 8 megapixels, analyzed using OpenCV library, and the second income (clothing type), determined by the local binary pattern algorithm, which is common digital image processing algorithm that widely used to identify shapes that follow specific pattern and structure, the third income (degree of dirt), and was identified by taking a picture of the clothes after soaking them with water for two minutes. The image was then analyzed by the OpenCV library and the fourth (washing weight) that getting From the Load Cell, which measures the physical weights. The readings were converted to digital values via the HX711 digital analogue converter and then sent to Arduino UNO to determine the weight. The weight values were eventually sent to the Raspberry PI for use in the controller. The system generates three exits: washing time (the length of time the laundry was washed), the temperature required for washing, and the amount of detergent required.
After selecting all the previous values, we transferred to control Wattar washing machine model 402, where the water valve was controlled to allow the water to pass into the powder box and from it to the washing basin. The water heater was controlled, which heated the water to the temperature determined by the Fuzzy algorithm, The temperature was monitored by the DS18B20 temperature sensor, which gives a signal to the Raspberry PI at the arrival of the temperature to the required value, and the washing machine engine is controlled for a third of the time specified in the Fuzzy algorithm and we controlled the pump Water to empty basin Washing from water, the process repeated for three consecutive times, we control using a software interface designed using TKinter library
We have been able to design a smart Fuzzy logic type-2 controller with the following advantages:
o save electricity consumption o Provide quantity of detergents o Shortenwashingtime
We have been able to control the following physical components within the washing machine:
o Control the water pump
o control Water valve
o controlMotor
o controlTemperaturesensor o controlLCDscreen
We have built a smart washing machine with the following characteristics:
o Have the ability to recognize the condition of clothes o Identify the type of clothing
o Identify the color of clothes
o Dothewashingwithoutusingapredefinedprogram.
The controller we designed gives good results to calculate the following:
o Washingtime
o Quantity of detergents o Temperature
All diagrams appear in the case of the incremental gradient with an increased degree of dirt and as values correspond to each type of clothing.
Keywords: smart washing machine, saving electricity, saving detergent, shortening washing time, color and clothing distinction, artificial intelligence, fuzzy logic type-2, Raspberry PI, control, Python programming language, HX711.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
