Browsing by Author "Choupani, Roya"

Now showing 1 - 12 of 12

Citation - WoS: 0
Citation - Scopus: 4
A drift-reduced scheme for hierarchical wavelet coding scalable video transmissions
(Ieee, 2009) Choupani, Roya; Wong, Stephan; Tolun, Mehmet R.
Scalable video coding allows for the capability of (partially) decoding a video bitstream when faced with communication deficiencies such as low handwidth or loss of data resulting in lower video quality. As the encoding is usually based on perfectly reconstructed frames, such deficiencies result in differently decoded frames at the decoder than the ones used in the encoder and, therefore, leading to errors being accumulated in the decoder. This is commonly referred to as the drift error. Drift-free scalable video coding methods also suffer from the low performance problem as they do not combine the residue encoding scheme of the current standards such as MPEG-4 and H.264 with scalability characteristics. We propose a scalable video coding method which is based on the motion compensation and residue encoding methods found in current video standards combined with the scalability property of discrete wavelet transform. Our proposed method aims to reduce the drift error while preserving the compression efficiency. Our results show that the drift error has been greatly reduced when a hierarchical structure for frame encoding is introduced.
Citation - WoS: 0
Citation - Scopus: 1
Adaptive Embedded Zero Tree For Scalable Video Coding
(int Assoc Engineers-iaeng, 2011) Choupani, Roya; Wong, Stephan; Tolun, Mehmet R.; 1863
Video streaming over the Internet has gained popularity during recent years mainly due to the revival of video-conferencing and video-telephony applications and the proliferation of (video) content providers. However, the heterogeneous, dynamic, and best-effort nature of the Internet cannot always guarantee a certain bandwidth for an application utilizing the Internet. Scalability has been introduced to deal with such issues (up to a certain point) by adapting the video quality with the available bandwidth. In addition, wavelet based scalability combined with representation methods such as embedded zero trees (EZWs) provides the possibility of reconstructing the video even when only the initial part of the streams have been received. EZW prioritizes the wavelet coefficients based on their energy content. Our experiments however, indicate that giving more priority to low frequency content improves the video quality at a specific bit rate. In this paper, we propose a method to improve on the compression rate of the EZW by prioritizing the coefficients by combining each frequency sub-band with its energy content. Initial experimental show that the first two layers of the generated EZW are about 22.6% more concise.
Client-server communication in remote control
(2002) Choupani, Roya
The client server communication model has been used in remote controlling of devices. The main feature in this study is that the Internet has been used as the common media to transmit controlling data and receive information. Client server model on the Internet restricts the access to client computers and has the disadvantage of unknown platform on the client side. This problem has been solved by means of platform independent programming and Java applets. Socket interface available in application layer of TCP/IP protocol suit has been used to establish reliable connection between clients and server. Security issues have been dealt with in the server side by checking the IP addresses of requesting clients.
Face Photograph Recognition via Generation from Sketches using Convolutional Neural Networks
(2019) Karasolak, Mustafa; Choupani, Roya; 21259
Face photo-sketch matching is an important problem for law enforcement agencies in terms of identifying suspects. In this study, a new sketch-photo generation and recognition technique is proposed by using residual convolutional neural network architecture. The suggested RCNN architecture consists of 6 convolutions, 6 ReLU, 4 poolings, 2 deconvolution layers. The proposed architecture is trained with face photos and sketches. Sketches are supplied as an input to the RCNN architecture and, generated face photos are obtained as the output. Then, the generated face photos are compared with the photos of the people in the database. Structural Similarity Index (SSIM) is used to measure the pairwise similarity and the photo with the highest index score is matched. CUHK Face Sketch Database containing 188 images is tested. In the experiments, 148, 20, and 20 images are used for training, validation, and testing, respectively. Data augmentation applied to 148 training images produced 444 images. Experimental results show that the success of the training curve is 90.55% and the validation success is 91.1%. True face recognition success from generated face images with SSIM is 93.89% for CUHK Face Sketch database (CUFS) and 84.55% AR database.
Hierarchical SNR Scalable Video Coding with Adaptive Quantization for Reduced Drift Error
(2015) Choupani, Roya; Wong, Stephan; Tolun, Mehmet
In video coding, dependencies between frames are being exploited to achieve compression by only coding the differences. This dependency can potentially lead to decoding inaccuracies when there is a communication error, or a deliberate quality reduction due to reduced network or receiver capabilities. The dependency can start at the reference frame and progress through a chain of dependent frames within a group of pictures (GOP) resulting in the so-called drift error. Scalable video coding schemes should deal with such drift errors while maximizing the delivered video quality. In this paper, we present a multi-layer hierarchical structure for scalable video coding capable of reducing the drift error. Moreover, we propose an optimization to adaptively determine the quantization step size for the base and enhancement layers. In addition, we address the trade-off between the drift error and the coding efficiency. The improvements in terms of average PSNR values when one frame in a GOP is lost are 3.70(dB) when only the base layer is delivered, and 4.78(dB) when both the base and the enhancement layers are delivered. The improvements in presence of burst errors are 3.52(dB) when only the base layer is delivered, and 4.50(dB) when both base and enhancement layers are delivered.
Mugshot Matching via Generation from Sketches using Convolutional Neural Networks
(2019) Choupani, Roya; 21259
Citation - WoS: 5
Citation - Scopus: 13
Multi-Label Classification of Text Documents Using Deep Learning
(Ieee, 2020) Mohammed, Hamza Haruna; Dogdu, Erdogan; Gorur, Abdul Kadir; Choupani, Roya
Recently, studies in the field of Natural Language Processing and its related applications continue to mount up. Machine learning is proven to be predominantly data-driven in the sense that generic model building methods are used and then tailored to specific application domains. Needless to say, this has proven to be a very effective approach in modeling the complicated data dependencies we frequently experience in practice, making very few assumptions, and allowing the information to talk for themselves. Examples of these applications can be found in chemical process engineering, climate science, healthcare, and linguistic processing systems for natural languages, to name a few. Text classification is one of the important machine learning tasks that is used in many digital applications today; such as in document filtering, search engines, document management systems, and many more. Text classification is the process of categorizing of text documents into a given set of labels. Furthermore, multi-label text classification is the task of categorization of text documents into one or more labels simultaneously. Over the years, many methods for classifying text documents have been proposed, including the popularly known bag of words (BoW) method, support vector machine (SVM), tree induction, and label-vector embedding, to mention a few. These kinds of tools can be used in many digital applications, such as document filtering, search engines, document management systems, etc. Lately, deep learning-based approaches are getting more attention, especially in extreme multi-label text classification case. Deep learning has proven to be one of the major solutions to many machine learning applications, especially those involving high-dimensional and unstructured data. However, it is of paramount importance in many applications to be able to reason accurately about the uncertainties associated with the predictions of the models. In this paper, we explore and compare the recent deep learning-based methods for multi-label text classification. We investigate two scenarios. First, multi-label classification model with ordinary embedding layer, and second with Glove, word2vec, and FastText as pre-trained embedding corpus for the given models. We evaluated these different neural network model performances in terms of multi-label evaluation metrics for the two approaches, and compare the results with the previous studies.
Citation - WoS: 4
Citation - Scopus: 6
Multiple description coding for SNR scalable video transmission over unreliable networks
(Springer, 2014) Choupani, Roya; Wong, Stephan; Tolun, Mehmet
Streaming multimedia data on best-effort networks such as the Internet requires measures against bandwidth fluctuations and frame loss. Multiple Description Coding (MDC) methods are used to overcome the jitter and delay problems arising from frame losses by making the transmitted data more error resilient. Meanwhile, varying characteristics of receiving devices require adaptation of video data. Data transmission in multiple descriptions provides the feasibility of receiving it partially and hence having a scalable and adaptive video. In this paper, a new method based on integrating MDC and signal-to-noise ratio (SNR) scalable video coding algorithms is proposed. Our method introduces a transform on data to permit transmitting them using independent descriptions. Our results indicate that on average 1.71dB reduction in terms of Y-PSNR occurs if only one description is received.
Citation - WoS: 2
Multiple Description Scalable Coding for Video Transmission over Unreliable Networks
(Springer-verlag Berlin, 2009) Choupani, Roya; Wong, Stephan; Tolun, Mehmet R.; 1863
Developing real time multimedia applications for best effort networks such as the Internet requires prohibitions against jitter delay and frame loss. This problem is further complicated in wireless networks as the rate of frame corruption or loss is higher in wireless networks while they generally have lower data rates compared to wired networks. On the other hand, variations of the bandwidth and the receiving device characteristics require data rate adaptation capability of the coding method. Multiple Description Coding (MDC) methods are used to solve the jitter delay and frame loss problems by making the transmitted data more error resilient, however, this results in reduced data rate because of the added overhead. MDC methods do not address the bandwidth variation and receiver characteristics differences. In this paper a new method based on integrating MDC and the scalable video coding extension of H.264 standard is proposed. Our method can handle both jitter delay and frame loss, and data rate adaptation problems. Our method utilizes motion compensating scheme and, therefore, is compatible with the current video coding standards such as MPEG-4 and H.264. Based on the simulated network conditions, our method shows promising results and we have achieved tip to 36dB for average Y-PSNR.
Citation - WoS: 7
Phishing e-mail detection by using deep learning algorithms
(Assoc Computing Machinery, 2018) Hassanpour, Reza; Doğdu, Erdoğan; Hassanpour, Reza; Dogdu, Erdogan; Choupani, Roya; Goker, Onur; Nazli, Nazli; 21259; Yazılım Mühendisliği; Bilgisayar Mühendisliği
Citation - WoS: 1
Unbalanced Multiple Description Wavelet Coding for Scalable Video Transmission
(Spie-soc Photo-optical instrumentation Engineers, 2012) Choupani, Roya; Wong, Stephan; Tolun, Mehmet; 21259
Scalable video coding and multiple description coding are the two different adaptation schemes for video transmission over heterogeneous and best-effort networks such as the Internet. We propose a new method to encode video for unreliable networks with rate adaptation capability. Our proposed method groups three dimensional discrete wavelet transform coefficients in different descriptions and applies a modified embedded zero tree data for rate adaptation. The proposed method optimizes the bit-rates of the descriptions with respect to the channel bit rates and the maximum acceptable distortion. The experimental results in the presence of one description loss indicate that on average the videos at the rate of 1000 Kbit/s are reconstructed with Y-component of peak signal to noise ratio (Y-PSNR) value of 36.2 dB. The dynamic allocation of descriptions to the network channels is optimized for rate distortion minimization. The improvement in term of Y-PSNR achieved by rate distortion optimization has been between 0.7 and 5.3 dB in different bit rates. (c) 2012 SPIE and IS&T. [DOI: 10.1117/1.JEI.21.4.043006]
Using wavelet transform self-similarity for effective multiple description video coding
(IEEE, 2016) Choupani, Roya; Wong, Stephan; Tolun, Mehmet
Video streaming over unreliable networks requires preventive measures to avoid quality deterioration in the presence of packet losses. However, these measures result in redundancy in the transmitted data which is utilized to estimate the missing packets lost in the delivered portions. In this paper, we have used the self-similarity property if the discrete wavelet transform (DWT) to minimize the redundancy and improve the fidelity of the delivered video streams in presence of data loss. Our proposed method decomposes the video into multiple descriptions after applying the DWT. The descriptions are organized in such a way that when one of them is lost during transmission, it is estimated using the delivered portions by means of self-similarity between the DWT coefficients. In our experiments, we compare video reconstruction in the presence of data loss in one or two descriptions. Based on the experimental results, we have ascertained that our estimation method for missing coefficients by means of self-similarity is able to improve the video quality by 2.14dB and 7.26dB in case of one description and two descriptions, respectively. Moreover, our proposed method outperforms the state-of-the-art Forward Error Correction (FEC) method in case of higher bit-rates.