Anton Lebedev

Bio: Anton Lebedev is an academic researcher from Yaroslavl State University. The author has contributed to research in topics: Convolutional neural network & Face detection. The author has an hindex of 5, co-authored 11 publications receiving 51 citations.

Facial expression recognition algorithm based on deep convolution neural network

Abstract: This paper presents algorithms for smile detection and facial expression recognition. The developed algorithms are based on the implementation of a relatively new approach in the field of deep machine learning — a convolutional neural network. The aim of this network is to classify facial images into one of the six types of emotions. The studying of algorithms was carried using face images from the CMU MultiPie database. To accelerate the neural network operation, the training and testing processes were performed parallel, on a large number of independent streams on GPU. Fo r developed models there were given metrics of quality.

Age estimation from face images: challenging problem for audience measurement systems

Abstract: The real-time audience measurement system consists of five consecutive stages: face detection, face tracking, gender recognition, age classification and in-cloud data statistics analysis. The challenging part of such system is age estimation algorithm on the basis of machine learning methods. The face aging process is determined by different factors: genetic, lifestyle, expression and environment. That is why same age people can have quite different rates of facial aging. We propose a novel algorithm consisting of two stages: adaptive feature extraction based on local binary patterns and support vector machine classification. Experimental results on the FG-NET, MORPH and our own database are presented. Human perception ability in age estimation is studied using crowdsourcing which allows a comparison of the ability of machines and humans.

Using Convolutional Neural Networks in the Problem of Cell Nuclei Segmentation on Histological Images

Abstract: Computer-aided diagnostics of cancer pathologies based on histological image segmentation is a promising area in the field of computer vision and machine learning. To date, the successes of neural networks in image segmentation in a number of tasks are comparable to human results and can even exceed them. The paper presents a fast algorithm of histological image segmentation based on the convolutional neural network U-Net. Using this approach allows to get better results in the tasks of medical image segmentation. The developed algorithm based on neural network AlexNet was used for the creation of the automatic markup of the histological image database. The neural network algorithms were trained and tested on the NVIDIA DGX-1 supercomputer using histological images. The results of the research show that the fast algorithm based on neural network U-Net can be successfully used for the histological image segmentation in real medical practice, which is confirmed by the high level of similarity of the obtained markup with the expert one.

Evaluation of face image quality metrics in person identification problem

Abstract: Face quality assessment algorithms play an important role in improving face recognition accuracy and increasing computational efficiency of biometric systems. In the case of video analysis system, it is very common to acquire multiple face images of a single person. Strategy for optimally choose of the face images with the best quality from the set of images should base on special quality metric. A set of face image quality metrics were investigated: image resolution, sharpness, symmetry, blur, measure of symmetry of landmarks points S, quality measure based on learning to rank. A new metric based on no-reference image quality assessment approach is proposed. For all the metrics the Spearman rank correlation coefficients with subjective expert assessment at different levels of face image scene illumination were calculated. The received results can help computer vison system engineers to optimize the biometric identification system.

Automatic Identification of Appendiceal Orifice on Colonoscopy Images Using Deep Neural Network

Abstract: The results of testing the recognition algorithm of the cecum achievement in colonoscopy video of the colon mucosa are presented. The image database was formed from the results of colonoscopy procedure together with the doctors of the Yaroslavl Regional Oncology Hospital. As the architecture of the convolutional neural network, the ResNet50 modification, previously trained on the standard ImageNet base, was chosen. As a result of applying the machine learning algorithm to the test set of endoscopic images, the metric values were AUC=0.95, Fscore=0.9, when a threshold is h=0.462. The results can be used to develop a quality control system for colonoscopy procedures. The introduction of such a system in medical practice will partially automate the analysis of video data, which will subsequently lead to a decrease in the number of subjective medical errors during colonoscopy.

MORPH: A Longitudinal Image Database of Normal Adult Age-Progression.

Abstract: This paper details MORPH a longitudinal face database developed for researchers investigating all facets of adult age-progression, e.g. face modeling, photo-realistic animation, face recognition, etc. This database contributes to several active research areas, most notably face recognition, by providing: the largest set of publicly available longitudinal images; longitudinal spans from a few months to over twenty years; and, the inclusion of key physical parameters that affect aging appearance. The direct contribution of this data corpus for face recognition is highlighted in the evaluation of a standard face recognition algorithm, which illustrates the impact that age-progression, has on recognition rates. Assessment of the efficacy of this algorithm is evaluated against the variables of gender and racial origin. This work further concludes that the problem of age-progression on face recognition (FR) is not unique to the algorithm used in this work.

Facial Expression Recognition Using Facial Landmarks and Random Forest Classifier

Abstract: Human emotions are the universally common mode of interaction. Automated human facial expression identification has its own advantages. In this paper, the author has proposed and developed a methodology to identify facial emotions using facial landmarks and random forest classifier. Firstly, faces are identified in each image using a histogram of oriented gradients with a linear classifier, image pyramid, and sliding window detection scheme. Then facial landmarks are identified using a model trained with the iBUG 300-W dataset. A feature vector is calculated using a proposed method which uses identified facial landmarks and it is normalized using a proposed method in order to remove facial size variations. The same feature vector is calculated for the neutral pose and vector difference is used to identify emotions using random forest classifier. Famous Extended Cohn-Kanade database has been used to train random forest and to test the accuracy of the system.

Geometric-Convolutional Feature Fusion Based on Learning Propagation for Facial Expression Recognition

Abstract: Facial expression is the main approach for humans to express their emotions. It is the temporal-spatial information that can be recognized by computers. In this paper, three video-based models are proposed for the facial expression recognition system (FERS). First, a differential geometric fusion network (DGFN) is proposed, which utilizes the technique of the handcrafted feature for traditional machine learning. The static geometric feature in the DGFN, which is based on the critical regions of psychology and the rules of physiology, is converted into the differential geometric feature by the geometric fusion model. Then deep-facial-sequential network (DFSN) is designed based on a multi-dimensional convolutional neural network (CNN). Finally, the DFSN-I is proposed, which is the combination of the DGFN and the DFSN taking advantages of both to achieve better performance. The experimental result shows that the combination of the handcrafted feature with prior experience and the auto-extracted feature provides better performance. It also shows that our DFSN and DFSN-I outperform the state-of-the-art methods on the Oulu-CASIA data set and achieve almost the best performance on CK+ compared with the other video-based methods.

Standardization of Face Image Sample Quality

Abstract: Performance of biometric systems is dependent on quality of acquired biometric samples. Poor sample quality is a main reason for matching errors in biometric systems and may be the main weakness of some implementations. This paper proposes an approach for standardization of facial image quality, and develops facial symmetry based methods for the assessment of it by measuring facial asymmetries caused by non-frontal lighting and improper facial pose. Experimental results are provided to illustrate the concepts, definitions and effectiveness.