Araştırma Çıktıları | WoS | Scopus | TR-Dizin | PubMed
Permanent URI for this communityhttps://hdl.handle.net/20.500.14719/1741
Browse
25 results
Search Results
Publication Metadata only An hierarchical approach for human computer interaction using eyelid movements, Göz kapaǧi hareketleriyle insan bilgisayar etkileşimi için siradüzensel yaklaşim(IEEE Computer Society [email protected], 2014) Çelik, Anıl; Arica, Nafiz; Çelik, Anıl, Bahçeşehir Üniversitesi, Istanbul, Turkey; Arica, Nafiz, Bahçeçehir Üniversitesi, Istanbul, TurkeyThis work proposes a method to achieve real-time HumanComputer interaction with the movements of the eyelids in low-resolution video. Classification of left and right eye states as closed or open are performed by an hierarchical approach of tracking by detection. After the initial detection of the face area, an efficient face tracking algorithm is used to reduce the search space for detection of eye region. By seperating the eye region into two overlapping pieces, left and right eyes are detected and classified as closed or open. The proposed method is robust against mimics and multiple faces in the frame, while being unaffected by the negative effects of aliasing and resizing. Thus, people who suffer from medical conditions limiting their physical movement capabilities can interact with computers with their eyelid movements. © 2014 IEEE. © 2014 Elsevier B.V., All rights reserved.Publication Metadata only Atmospheric turbulence mitigation using feature based optical flow, Atmosferik türbülans etkilerinin öznitelik tabanli optik akiş yöntemi ile azaltilmasi(IEEE Computer Society [email protected], 2014) Çaliskan, Tufan; Arica, Nafiz; Çaliskan, Tufan, Deniz Harp Okulu, Tuzla, Turkey; Arica, Nafiz, Mühendislik Fakültesi, Bahçeşehir Üniversitesi, Istanbul, TurkeyIn this paper, we propose a fast and effective feature based registration technique in removing the distortions caused by atmospheric turbulence. We utilize optical flow method and combine the advantages of previous approaches based on space-invariant deconvolution and lucky frame idea. After an optical flow based registration of degraded image sequence, a patch-wise multi-frame reconstruction technique is applied to fuse the registered images. Blind-deconvolution technique is implemented to deblur the fused image to obtain a single high quality image. The experiments performed on common datasets show that the proposed method produces higher quality images more efficiently than the available methods. © 2014 IEEE. © 2014 Elsevier B.V., All rights reserved.Publication Metadata only Atmospheric turbulence mitigation using optical flow(Institute of Electrical and Electronics Engineers Inc., 2014) Çaliskan, Tufan; Arica, Nafiz; Çaliskan, Tufan, Department of Computer Engineering, Deniz Harp Okulu, Tuzla, Turkey; Arica, Nafiz, Faculty of Engineering, Bahçeşehir Üniversitesi, Istanbul, TurkeyAtmospheric turbulence causes blurring and geometrical distortion in images acquired from a long distance. In this paper, we propose a fast and effective feature based registration technique in removing the distortions caused by atmospheric turbulence. We utilize optical flow method and combine the advantages of previous approaches based on space invariant deconvolution and lucky frame idea. After an optical flow based registration of degraded image sequence, a patch-wise multi-frame reconstruction technique is applied to fuse the registered images. Finally, a blind-deconvolution technique is implemented to deblur the fused image to obtain a single high quality image. The experiments performed on common datasets show that the proposed method produces higher quality images more efficiently than the available methods. © 2021 Elsevier B.V., All rights reserved.Publication Metadata only Hierarchical image representation using deep network(Springer Verlag [email protected], 2015) Ergül, Emrah; Ertürk, Sarp; Arica, Nafiz; Murino, V.; Murino, V.; Puppo, E.; Ergül, Emrah, Kocaeli Üniversitesi, İzmit, Turkey; Ertürk, Sarp, Kocaeli Üniversitesi, İzmit, Turkey; Arica, Nafiz, Bahçeşehir Üniversitesi, Istanbul, TurkeyIn this paper, we propose a new method for features learning from unlabeled data. Basically, we simulate k-means algorithm in deep network architecture to achieve hierarchical Bag-of-Words (BoW) representations. We first learn visual words in each layer which are used to produce BoW feature vectors in the current input space. We transform the raw input data into new feature spaces in a convolutional manner such that more abstract visual words are extracted at each layer by implementing Expectation-Maximization (EM) algorithm. The network parameters are optimized as we keep the visual words fixed in the Expectation step while the visual words are updated with the current parameters of the network in the Maximization step. Besides, we embed spatial information into BoW representation by learning different networks and visual words for each quadrant regions. We compare the proposed algorithm with the similar approaches in the literature using a challenging 10-class-dataset, CIFAR-10. © 2015 Elsevier B.V., All rights reserved.Publication Metadata only Occlusion analysis for face frontalization(Institute of Electrical and Electronics Engineers Inc., 2016) Çelik, Ami; Arica, Nafiz; Bayrak, C.; Varol, C.; Ozturk, Y.; Çelik, Ami, Department of Computer Engineering, Bahçeşehir Üniversitesi, Istanbul, Turkey; Arica, Nafiz, Department of Computer Engineering, Bahçeşehir Üniversitesi, Istanbul, TurkeyFrontalization is the process of generating frontal faces from the posed ones appearing in unconstrained environments. Occluders occurring over the face region can make the frontalization approaches generate faulty results. In this study, we propose an approach, to address this problem by reducing negative effects of occluders on the frontalization operation. The proposed approach has the capability of choosing the most suitable frontalization procedure, to generate a more visually appealing output. After the posed face image is hard-frontalized, the possible occlusion occurrences over the face are analyzed using two different techniques, region and pixel based analysis. Finally, according to occlusion analysis output, the best approach for frontalization process is chosen to be applied The experiments performed on Caltech Occluded Faces Database (COFW) show that the proposed algorithm produces satisfactory results in terms of accuracy and visual appearance. © 2017 Elsevier B.V., All rights reserved.Publication Metadata only Object tracking by combining tracking-by-detection and marginal particle filter, Tespit-ile-Takip ve Marjinal Parçacik Süzgecinin Birleştirilmesiyle Nesne Takibi(Institute of Electrical and Electronics Engineers Inc., 2016) Maraş, Bahri; Arica, Nafiz; Ertüzün, Ayşin Baytan; Maraş, Bahri, Boğaziçi Üniversitesi, Bebek, Turkey; Arica, Nafiz, Bilgisayar Mühendisliǧi Bölümü, Bahçeşehir Üniversitesi, Istanbul, Turkey; Ertüzün, Ayşin Baytan, Boğaziçi Üniversitesi, Bebek, TurkeyIn this paper, we propose a new algorithm based on tracking-by-detection approach for tracking of the objects having non-linear dynamic motion. For this purpose, the tracking-by-detection method depending on the Gauss kernel using the circulant matrix theory and the Fourier transform is employed together with the marginal particle filter method. Marginal particle filter uses the scores derived from the Gauss kernel at the measurement update phase of the filter to weight the particles propagated around the target that has been tracked. While updating the state variables by marginal particle filter, the object coordinates, the correction values belonging to these coordinates and the dimensions of the image window surrounding the object is estimated. The proposed method is tested on the video sequences which include the object having a high non-linear motion, and it was observed that marginal particle filter enhanced the performance of the track-by-detection method in an important scale. © 2017 Elsevier B.V., All rights reserved.Publication Metadata only Risk classification for breast cancer diagnosis using HER2 testing, Meme Kanseri Tanisi için HER2 Testine Dayali Risk Siniflandirmasi(Institute of Electrical and Electronics Engineers Inc., 2016) Aktan, Pinar Ece; Hatipoglu, Gokhan; Arica, Nafiz; Aktan, Pinar Ece, Bilgisayar Mühendisliǧi Bölümü, Bahçeşehir Üniversitesi, Istanbul, Turkey; Hatipoglu, Gokhan, Virasoft, Istanbul, Turkey; Arica, Nafiz, Bilgisayar Mühendisliǧi Bölümü, Bahçeşehir Üniversitesi, Istanbul, TurkeyTrastuzumab is one of the effective treatment for breast cancer. It is an antibody against the protein product of the human epidermal growth factor receptor type 2 (HER2) gene. In HER-2 testing protein levels are determined by IHC staining. For a proper treatment IHC stains are evaluated together with ISH stains. The widespread approach done by the experts is to determine membranes manually, yet it leads subjective interpretation by pathologist and it increases their workload. In this paper, we propose a new method for HER-2 testing based on image analysis algorithms on membrane regions. After, maximally stable extremal regions algorithm is employed for detecting the membrane regions, these areas of those regions are analyzed quantitatively for IHC classification. In the next step, complete/incomplete membrane regions are determined for IHC 2+ classification. The experiments conducted on the IHC stains taken from the clinic patients. Analysis resulted with the successful classification rates. © 2017 Elsevier B.V., All rights reserved.Publication Metadata only Size variant landmark patches for Facial Action Unit detection(Institute of Electrical and Electronics Engineers Inc., 2016) Çakir, Duygu; Arica, Nafiz; Saha, H.N.; Chakrabarti, S.; Çakir, Duygu, Faculty of Engineering and Natural Sciences, Bahçeşehir Üniversitesi, Istanbul, Turkey; Arica, Nafiz, Faculty of Engineering and Natural Sciences, Bahçeşehir Üniversitesi, Istanbul, TurkeyFacial Action Coding System (FACS) provides the description of all visual facial changes in terms of Action Units (AU), which express the movements of facial muscles. In this study we introduce a sparse learning method for AU detection based on facial landmark patches. The proposed method learns the active landmarks for each AU while finding their best representative patch sizes in a unified framework. The experiments performed on the publicly available CK+ dataset show that the proposed approach outperforms the state-of-the-art studies by gaining the highest F1 scores for almost all of the lower face Action Units. The experiments also show that the subtle muscle movements belonging to the upper face require smaller landmark patches while the lower face AUs are detected better in larger patches. © 2017 Elsevier B.V., All rights reserved.Publication Metadata only Face frontalization enhanced by deep learning, Derin Öǧrenme Destekli Yüz Önleştirme(Institute of Electrical and Electronics Engineers Inc., 2017) Çelik, Anıl; Arica, Nafiz; Çelik, Anıl, Bilgisayar Mühendisliǧi Bölümü, Bahçeşehir Üniversitesi, Istanbul, Turkey; Arica, Nafiz, Bilgisayar Mühendisliǧi Bölümü, Bahçeşehir Üniversitesi, Istanbul, TurkeyIn this study, a new approach based on 3-D models and deep learning to frontalize face images is proposed. Specifically designed for facial expression analysis, the proposed approach aims to reduce possible negative effects that a posed face image can generate, by normalizing the face region. In the first phase, the face image is semi-frontalized, with a pre-established 3-D reference model based approach. Then, missing regions on semi-frontalized images due to geometric transformation are reconstructed with the help of a denoising stacked autoencoder network. In this phase, missing regions created by line of sight are learned, with a deep architecture, using numerous images. When examined, it can be said that, faces acquired with the proposed approach, are objectively better than the faces acquired with a deep learning or 3-D based method alone. Therefore, it is assumed that the proposed approach can be used in face based computer vision methods as a beneficial pre-processing step. © 2017 Elsevier B.V., All rights reserved.Publication Metadata only Facial emotion classification using deep embedding with triplet loss function, Üçlü Yitim Fonksiyonu Tabanli Derin Gömme ile YÜz Duygu Siniflandirmasi(Institute of Electrical and Electronics Engineers Inc., 2017) Bircanoglu, Cenk; Arica, Nafiz; Bircanoglu, Cenk, Bilgisayar Mühendisliǧi Bölümü, Bahçeşehir Üniversitesi, Istanbul, Turkey; Arica, Nafiz, Bilgisayar Mühendisliǧi Bölümü, Bahçeşehir Üniversitesi, Istanbul, TurkeyIn this paper, a deep embedding method using triplet loss function is proposed for classification of the emotions in face images. The originality of proposed method lies in the loss function, different from the other deep learning-based facial emotion classification approaches. The input face images are embedded into a lower dimensional feature space using a multi-layer convolutional neural network. In this embedding process the loss function is calculated by taking triple samples in the training dataset. For each sample in batches, two samples are selected in such a way that one of them is from the same class and the other one from the different class. The loss function aims to close the samples belonging to the same class to each other in the Euclidean space and to move away from the samples in the different class. In the performance analysis of proposed method, two popular deep architectures, namely AlexNet and VGG are used on two datasets called CIFE and GaMo. The experiments show that the proposed method outperforms the other popular loss functions and the available studies on this particular datasets. © 2017 Elsevier B.V., All rights reserved.
- «
- 1 (current)
- 2
- 3
- »