In-plane Rotational Alignment of Faces by Eye and Eye-pair Detection

Karaaba, Mahir Faik; Surinta, Olarik; Schomaker, Lambert; Wiering, Marco A.

doi:10.5220/0005308303920399

Cited by 4 publications

(3 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this method, eye centers are used to compute the roll angle of the face. Then the face is rotated to roll-normalized position as described in (Karaaba et al, 2015). All eye coordinates are obtained from the dataset directories, except for some images (of each subject) of the FERET dataset for which we used an automatic alignment algorithm.…”

Section: Alignmentmentioning

confidence: 99%

Robust Face Identification with Small Sample Sizes using Bag of Words and Histogram of Oriented Gradients

Karaaba

Surinta

Schomaker

et al. 2016

Proceedings of the 11th Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications

Self Cite

View full text Add to dashboard Cite

Face identification under small sample conditions is currently an active research area. In a case of very few reference samples, optimally exploiting the training data to make a model which has a low generalization error is an important challenge to create a robust face identification algorithm. In this paper we propose to combine the histogram of oriented gradients (HOG) and the bag of words (BOW) approach to use few training examples for robust face identification. In this HOG-BOW method, from every image many sub-images are first randomly cropped and given to the HOG feature extractor to compute many different feature vectors. Then these feature vectors are given to a K-means clustering algorithm to compute the centroids which serve as a codebook. This codebook is used by a sliding window to compute feature vectors for all training and test images. Finally, the feature vectors are fed into an L2 support vector machine to learn a linear model that will classify the test images. To show the efficiency of our method, we also experimented with two other feature extraction algorithms: HOG and the scale invariant feature transform (SIFT). All methods are compared on two well-known face image datasets with one to three training examples per person. The experimental results show that the HOG-BOW algorithm clearly outperforms the other methods.

show abstract

Section: Alignmentmentioning

confidence: 99%

Robust Face Identification with Small Sample Sizes using Bag of Words and Histogram of Oriented Gradients

Karaaba

Surinta

Schomaker

et al. 2016

Proceedings of the 11th Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications

Self Cite

View full text Add to dashboard Cite

show abstract

“…To obtain the eye centers, we used the manual crop information provided in the dataset folder, except for the FERET dataset, from which we cropped the face images automatically by our eye and eye-pair detector (since FERET does not provide sufficient ground truth information for each image) and replaced badly cropped ones with manual crops (around 5% of them). After obtaining eye-coordinates, we followed the aligning method as presented in [34].…”

Section: ) Test Datasetsmentioning

confidence: 99%

Robust Face Recognition by Computing Distances From Multiple Histograms of Oriented Gradients

Karaaba

Surinta

Schomaker

et al. 2015

2015 IEEE Symposium Series on Computational Intelligence

View full text Add to dashboard Cite

The Single Sample per Person Problem is a challenging problem for face recognition algorithms. Patch-based methods have obtained some promising results for this problem. In this paper, we propose a new face recognition algorithm that is based on a combination of different histograms of oriented gradients (HOG) which we call Multi-HOG. Each member of Multi-HOG is a HOG patch that belongs to a grid structure. To recognize faces, we create a vector of distances computed by comparing train and test face images. After this, a distance calculation method is employed to calculate the final distance value between a test and a reference image. We describe here two distance calculation methods: mean of minimum distances (MMD) and a multi-layer perceptron based distance (MLPD) method. To cope with aligning difficulties, we also propose another technique that finds the most similar regions for two compared images. We call it the most similar region selection algorithm (MSRS). The regions found by MSRS are given to the algorithms we proposed. Our results show that, while MMD and MLPD contribute to obtaining much higher accuracies than the use of a single histogram of oriented gradients, combining them with the most similar region selection algorithm results in state-of-the-art performances.

show abstract

“…For solution of this issue, we can rotate whole face by calculating distance from X and Y axis of both eyes. y = left eye from Y axis − right eye from Y axis x = left eye from X axisright eye from X axis Rotation angle = arctan (y/x) [7,8] Using above equation we have rotate whole face to prepare it straight then send for further steps. End of this step all faces are extracted and store it in form of feature vector for feature extraction.…”

Section: Face Features Vector Generationmentioning

confidence: 99%

Human recognition from multi angular images

Naik¹,

Lad²

Kalpa Publications in Computing

View full text Add to dashboard Cite

Face recognition is still complicated task because to envision human actions might not realizable in each incident. Intention of face recognition is to identify human based on face that is similar from available dataset face images. Human face has multidimensional structure so it requires efficient technique for face harmonization and verification.Proposed work aim for developing efficient human face recognition method that deals with front as well as side view face in normal face expression. Using Viola–Jones face detection algorithm it accumulate only face region. Face features like eyes, nose and lip are extracted from whole face region using canny edge detection and harris corner detection method. To match individual face features, it compares position of edge boundary of features between images. Authors’ uses Euclidian distance method to retrieves maximum match value among all store face images. Based on threshold value it decides whether human face is recognized or not. Authors have evaluated performance of proposed method with DCT, DWD, PCA and LFL method on public free database like FEI, CVL and MIT-CBCL.

show abstract

In-plane Rotational Alignment of Faces by Eye and Eye-pair Detection

Cited by 4 publications

References 18 publications

Robust Face Identification with Small Sample Sizes using Bag of Words and Histogram of Oriented Gradients

Robust Face Identification with Small Sample Sizes using Bag of Words and Histogram of Oriented Gradients

Robust Face Recognition by Computing Distances From Multiple Histograms of Oriented Gradients

Human recognition from multi angular images

Contact Info

Product

Resources

About