Facial Features Analysis Using Deep Learning
As wrinkles are the most important features of aging, the anti-aging market to treat facial wrinkles is growing rapidly. A precise image-based analysis of those features can play a crucial role in relevant aging applications. In market cosmetic firms produces numerous anti-wrinkle creams, and dermat
2025-06-28 16:32:32 - Adil Khan
Facial Features Analysis Using Deep Learning
Project Area of Specialization Artificial IntelligenceProject SummaryAs wrinkles are the most important features of aging, the anti-aging market to treat facial wrinkles is growing rapidly. A precise image-based analysis of those features can play a crucial role in relevant aging applications. In market cosmetic firms produces numerous anti-wrinkle creams, and dermatological firms invest in wrinkle filler injections. While the wrinkles are easily distinguishable by the human eyes, it is a challenging task for machine learning to detect them automatically, as there’s no specific work and application in this scenario, especially in our country. In this project, we are introducing a deep learning based application which detects the wrinkles on your face with the help of convolution neural network, CNN model VGG and face recognition mechanism which give the results in percentage on the basis of training and comparisons. This application also compares the previous and current results with the help of taking good quality images by high resolution external camera on our system and will show the information that how many percent your wrinkles have been decreased or increased.
Project Objectives- To detect wrinkles on your face by using deep learning.
- To create an android application and system software to detect wrinkles on uploaded facial image by high resolution specific camera with good lightning.
- To evaluate the beneficial effects of treatments regarding wrinkles by comparing previous and current results with user friendly display methods.
- Our project is based on Machine Learning. In our project we are training our data set and then system predicts the results. Our first step is to make ready our system. First installed different libraries which are used in our model. The most important libraries are PANDAS, KERAS, OPEN CV, NUMPY, MATPLOT LIB.PYPLOT.
- We gathered data from different websites and data sets (i.e. HELEN, UTKFACES etc.) to make our data set of wrinkled and unwrinkled images. This is the most important part of our project because no labeled data set is available of wrinkled or unwrinkled images.
- After collecting data, we classified them into three parts training, validation and testing.
- After collecting data, we classified them into three parts training, validation and testing.
- After collecting data, we classified them into three parts training, validation and testing.
- After training our data using convolutional neural network we tested our data using testing data. We have to give one test image and our system gives result that our system is wrinkled or unwrinkled by using SOFTMAX activation function.
- The input test image is the on the spot uploaded image by high resolution camera with specific intensity of light for the accurate result because the lightning issue and camera result difference might give inaccurate result.
- So we need to use external hardware for static quality of result with easy and good display techniques for our customers for his past, present wrinkles estimation by comparisons.
It is a useful program for the medical profession, especially in facial skin line, and for the audience who are conscious about their skin, wrinkles, age and other face spots and their home treatments. The program will useful with them to analyze the methods to cure their problem by using the ground truth image from program that show the lines of wrinkles and spots on face. This can decrease the problem that nowadays we lack of doctors and they not have a chance to discuss or analyze the patient's problems. Doctors can use the ground truth image from program to follow up the change of the patient's problem in every period of times, they can know that the problems are better or not by compare the ground truth images to see the changes of the wrinkles. In market cosmetic companies produces various anti-wrinkle creams, and dermatological companies invest in wrinkle filler injections. we are introducing an application which detects the wrinkles and spots on your face and give the results in percentage at your home or detect by the doctors through this software.
Technical Details of Final Deliverable- The trained model to predict the wrinkleness of face.
- An android application which take an image as an input.
- An instruction manual for using the model and application.
| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 75000 | |||
| high resolution camera | Equipment | 1 | 15000 | 15000 |
| NVIDIA Graphic Card for training | Equipment | 1 | 55000 | 55000 |
| printing | Miscellaneous | 1 | 5000 | 5000 |