Partial Face Recognition
Vssen
I'm a undergradute
I need to conduct research and develop an artifact for partial face
Recogniton
it is a 8 month project
can you guys suggest me some readings please
Thanks
Viduruvan
PRATIK DAS
--
You received this message because you are subscribed to the Google Groups "Face Recognition Research Community" group.
To post to this group, send email to face...@googlegroups.com.
To unsubscribe from this group, send email to face-rec+u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/face-rec?hl=en.
ahmedgamal
me too work with this project
but from 4 months
at first you must work in face detection algorithms
like
1-knowledge based method
2-feature invariant method
3-template matching method
4-appearance based method
5-haar like method
6-generalized learning vector quantization
7-view based method
i working in my project
with C# &&MAT LAP
both parallel
tell me if you need more help
Kourosh Meshgi
Your categorization of face detection algorithms is kinda new to me, I wonder if you did this based on a published survey or it's your own ontology of problem? I also would like to ask you to share that survey to me if you had one.
Regards
Kourosh
PS: Do you work over partial occlusion or your work support self-occlusion too?
valentin dorneanu
Does anybody here work with Processing -:
I used it for face detection (very easy way)
and I wonder if
for face recognition could work as easy.
GoGo
glasses,accessories,etc
On Dec 25 2009, 9:59 pm, PRATIK DAS <jeetb...@gmail.com> wrote:
> what do you mean by partial face recognition??? what does the project
> actually do???
> please specify
>
>
>
> On Thu, Dec 24, 2009 at 8:39 PM, Vssen <vssen2...@gmail.com> wrote:
> > Hi
> > I'm a undergradute
>
> > I need to conduct research and develop an artifact for partial face
> > Recogniton
> > it is a 8 month project
>
> > can you guys suggest me some readings please
>
> > Thanks
>
> > Viduruvan
>
> > --
>
> > You received this message because you are subscribed to the Google Groups
> > "Face Recognition Research Community" group.
> > To post to this group, send email to face...@googlegroups.com.
> > To unsubscribe from this group, send email to
> > face-rec+u...@googlegroups.com<face-rec%2Bunsubscribe@googlegroups.com>
> > .
> > For more options, visit this group at
> >http://groups.google.com/group/face-rec?hl=en.- Hide quoted text -
>
> - Show quoted text -
Viduruvan S Godavaththa
Partial face recognition using core features of the face
Please describe the type of the system being developed
This system allows users to input (edited/non edited) specific regions of a face (mouth/nose and eye). Then the system will process the submitted image and identify features of the uploaded image. After that, by using characteristics of that uploaded partial face it will match the frontal (0 degree camera angle) face images which are already stored in the database and based on results it will recognise the face. The proposed artefact will develop a standalone application which runs on PC.
Why is such a system required?
Face is a unique and accurate biometric measure compared to other measures that are used to identify a specific person. The advantage of face recognition as a biometric identification technique is, it does not require people’s cooperation because people do not need to look into an iris scanner, to place their hands on a fingerprint reader, or to speak to a close-by microphone. Hence face recognition can be a useful footage taken in surveillance and security applications.
But when considering images taken from surveillance and security systems, it might not be possible to capture full frontal picture of a face at all times in uncontrolled environments. Even though there are many face recognition systems available, most of these work in optimal conditions. Especially if there does not exists full frontal face, these systems will fail to recognise a face. As a result of that most of system cannot give accurate face match results. Because of that there can be lot of complications identifying a person in an image.
By conducting a proper literature research and develop an artefact based on finding of the research I will try to overcome problems in recognition of faces using frontal partial face core features.
What will be the main challenge for you in building the system?
The main challenge of this project will be matching specific face regions with complete front faces stored in the database. The project challenge can be divided into the following sub sections
Pre processing uploaded image
The input image can be consisting of unwanted parts that are not required to identify the match. Therefore first of all, unwanted parts of the given image should be filtered out and only the required parts should be located.
Feature extraction
Identify unique characteristics in given partial facial image and by doing that identify core features in given image before processing those characteristic to match the faces.
Identify, locate and focus on a particular area of the face in the database to match.
After processing the submitted partial face, the particular face area in faces in the database to match should be identified.
Face Recognition based on submitted face features
Based on the characteristics and values given from the above process, the values with the faces in the database will be matched and results matched.
To overcome above sub challenges, it is necessary to carry out a good research to identify efficient, appropriate techniques and methods.
Currently for full frontal face recognition there are lot approaches to do that but when it come to partial face recognition it is very difficult to do face match based on separate features of a face. Finding an efficient approach will be challengeable thing in this project
What new ideas and theory will you need to learn to build the system?
There are different approaches to overcome above mentioned challenges. During the research stage of the project, the following theories and concepts will be studied. The research will carry out to study and identify efficient method to overcome above challenges.
Image processing (Image pre-processing for face recognition)
As mentioned before the submitted image should be processed based on a standard format which takes the system to match. Crop images, remove the noise from the image, binarising and convert images to required file format might be needed.
Face locating and detecting
Identifying face and other features should be done. This can be done using different techniques. Pattern matching and neural network basic methods can be identified as main methods.
Feature extraction and face identification using core features of the face
The system should be capable of identifying core features of the given face images and images in the database. This also can be done by using various techniques and algorithms.
Face recognition
Matching two faces can be done by using techniques but in this case it uses core features of the face to identify. Finding a method for this will be the main challenge of this project. Neural network basis methods and template matching techniques are the most common methods in the field. Learning methods and concepts including above methods should be done.
Database management with images
The images will be stored in a database therefore study about database should be done to improve efficiency of image storing and retrieving form the database which will directly affect to the efficiency of the system.
Vssen
> > Viduruvan- Hide quoted text -
Prashant Kumar
i am a b.tech student and doing my project on face recognition using
Gobar Wavelet Transformation and Cosine Transformation. Can you guys
please help in anyway?
Omid
The best thing to start your project is to find out what actually
Gabor WT or DCT are (easily search on google also wikipedia
could be helpful), then you could find how it really works on
face recognition,
rgds
> --
> You received this message because you are subscribed to the Google Groups
> "Face Recognition Research Community" group.
> To post to this group, send email to face...@googlegroups.com.
> To unsubscribe from this group, send email to
> face-rec+u...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/face-rec?hl=en.
>
>
--
Seyed Omid Shahdi
PhD Candidate
Computer Vision, Video, and Image Processing Research Lab
Dept. of Microelectronics and Computer Engineering
Faculty of Electrical Engineering
Universiti Teknologi Malaysia
81310 Skudai,Johor
Malaysia
email: osse...@siswa.utm.my
phone:+60177514001
Tommy Chai
Mind to share?
--
You received this message because you are subscribed to the Google Groups "Face Recognition Research Community" group.
To post to this group, send email to face...@googlegroups.com.
To unsubscribe from this group, send email to face-rec+u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/face-rec?hl=en.
--
Best regards,
Tommy Chai