Araştırma Çıktıları | WoS | Scopus | TR-Dizin | PubMed
Permanent URI for this communityhttps://hdl.handle.net/20.500.14719/1741
Browse
17 results
Search Results
Publication Metadata only ARVA: An Augmented Reality-Based Visual Aid for Mobility Enhancement Through Real-Time Video Stream Transformation(IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC, 2024) Sadeghzadeh, Arezoo; Islam, Md Baharul; Uddin, Md Nur; Aydin, Tarkan; Bahcesehir University; State University System of Florida; Florida Gulf Coast University; Daffodil International UniversityVisual field loss (VFL) is a persistent visual impairment characterized by limited vision spots (scotoma) within the normal visual field, significantly impacting daily activities for affected individuals. Current Virtual Reality (VR) and Augmented Reality (AR)-based visual aids suffer from low video quality, content loss, high levels of contradiction, and limited mobility assessment. To address these issues, we propose an innovative vision aid utilizing AR headset and integrating advanced video processing techniques to elevate the visual perception of individuals with moderate to severe VFL to levels comparable to those with unimpaired vision. Our approach introduces a pioneering optimal video remapping function tailored to the characteristics of AR glasses. This function strategically maps the content of live video captures to the largest intact region of the visual field map, preserving quality while minimizing blurriness and content distortion. To evaluate the performance of our proposed method, a comprehensive empirical user study is conducted including object counting and multi-tasking walking track tests and involving 15 subjects with artificially induced scotomas in their normal visual fields. The proposed vision aid achieves 41.56% enhancement (from 57.31% to 98.87%) in the mean value of the average object recognition rates for all subjects in object counting test. In walking track test, the average mean scores for obstacle avoidance, detected signs, recognized signs, and grasped objects are significantly enhanced after applying the remapping function, with improvements of 7.56% (91.10% to 98.66%), 51.81% (44.85% to 96.66%), 49.31% (43.18% to 92.49%), and 77.77% (13.33% to 91.10%), respectively. Statistical analysis of data before and after applying the remapping function demonstrates the promising performance of our method in enhancing visual awareness and mobility for individuals with VFL.Publication Metadata only Assistive Visual Tool: Enhancing Safe Navigation with Video Remapping in AR Headsets(SPRINGER INTERNATIONAL PUBLISHING AG, 2025) Sadeghzadeh, Arezoo; Islam, Md Baharul; Uddin, Md Nur; Aydin, Tarkan; DelBue, A; Canton, C; Pont-Tuset, J; Tommasi, T; Bahcesehir University; State University System of Florida; Florida Gulf Coast UniversityVisual Field Loss (VFL) is characterized by blind spots or scotomas that poses detrimental impact on fundamental movement activities of individuals. Addressing the challenges (e.g., low video quality, content loss, high levels of contradiction, and limited mobility assessment) faced by existing Extended Reality (XR) systems as vision aids, we introduce a groundbreaking method that enriches the real-time navigation using Augmented Reality (AR) glasses. Our novel vision aid employs advanced video processing techniques to enhance visual perception in individuals with moderate to severe VFL, bridging the gap to healthy vision. A unique optimal video remapping function, tailored to our selected AR glasses characteristics, dynamically maps live video content to the largest intact region of the Visual Field (VF) map. Our method preserves video quality, minimizing blurriness and distortion. Through a comprehensive empirical user study involving 29 subjects with artificially induced scotomas, statistical analyses of object counting and multi-tasking walking track tests demonstrate the promising performance of our method in enhancing visual awareness and navigation capability in real-time.Publication Metadata only HiMODE: A Hybrid Monocular Omnidirectional Depth Estimation Model(IEEE, 2022) Junayed, Masum Shah; Sadeghzadeh, Arezoo; Islam, Md Baharul; Wong, Lai-Kuan; Aydin, Tarkan; Bahcesehir University; Multimedia UniversityMonocular omnidirectional depth estimation is receiving considerable research attention due to its broad applications for sensing 360 degrees surroundings. Existing approaches in this field suffer from limitations in recovering small object details and data lost during the ground-truth depth map acquisition. In this paper, a novel monocular omnidirectional depth estimation model, namely HiMODE is proposed based on a hybrid CNN+Transformer (encoder-decoder) architecture whose modules are efficiently designed to mitigate distortion and computational cost, without performance degradation. Firstly, we design a feature pyramid network based on the HNet block to extract high-resolution features near the edges. The performance is further improved, benefiting from a self and cross attention layer and spatial/temporal patches in the Transformer encoder and decoder, respectively. Besides, a spatial residual block is employed to reduce the number of parameters. By jointly passing the deep features extracted from an input image at each backbone block, along with the raw depth maps predicted by the transformer encoder-decoder, through a context adjustment layer, our model can produce resulting depth maps with better visual quality than the ground-truth. Comprehensive ablation studies demonstrate the significance of each individual module. Extensive experiments conducted on three datasets, Stanford3D, Matterport3D, and SunCG, demonstrate that HiMODE can achieve state-of-the-art performance for 360 degrees monocular depth estimation. Complete project code and supplementary materials are available at https://github.com/himode5008/HiMODE.Publication Metadata only WNet: A dual-encoded multi-human parsing network(WILEY, 2024) Hosen, Md Imran; Aydin, Tarkan; Islam, Md Baharul; Bahcesehir UniversityIn recent years, multi-human parsing has become a focal point in research, yet prevailing methods often rely on intermediate stages and lacking pixel-level analysis. Moreover, their high computational demands limit real-world efficiency. To address these challenges and enable real-time performance, low-latency end-to-end network is proposed. This approach leverages vision transformer and convolutional neural network in a dual-encoded network, featuring a lightweight Transformer-based vision encoder) and a convolution encoder based on Darknet. This combination adeptly captures long-range dependencies and spatial relationships. Incorporating a fuse block enables the seamless merging of features from the encoders. Residual connections in the decoder design amplify information flow. Experimental validation on crowd instance-level human parsing and look into person datasets showcases the WNet's effectiveness, achieving high-speed multi-human parsing at 26.7 frames per second. Ablation studies further underscore WNet's capabilities, emphasizing its efficiency and accuracy in complex multi-human parsing tasks. We present WNet, a low-latency end-to-end network for multi-human parsing that integrates vision transformer and Convolutional Neural Network in a dual-encoded structure (vision encoder and a convolution encoder). By adeptly capturing long-range dependencies and spatial relationships, WNet achieves real-time performance and high-speed parsing at 26.7 frames per second on crowd instance-level human parsing and look into person datasets. The inclusion of a fuse block for seamless feature merging, along with residual connections in the decoder, amplifies information flow, emphasizing WNet's efficiency and accuracy in complex multi-human parsing tasks. imagePublication Metadata only Stereoscopic video quality measurement with fine-tuning 3D ResNets(SPRINGER, 2022) Imani, Hassan; Islam, Md Baharul; Junayed, Masum Shah; Aydin, Tarkan; Arica, Nafiz; Bahcesehir University; Daffodil International UniversityRecently, Convolutional Neural Networks with 3D kernels (3D CNNs) have shown great superiority over 2D CNNs for video processing applications. In the field of Stereoscopic Video Quality Assessment (SVQA), 3D CNNs are utilized to extract the spatio-temporal features from the stereoscopic video. Besides, the emergence of substantial video datasets such as Kinetics has made it possible to use pre-trained 3D CNNs in other video-related fields. In this paper, we fine-tune 3D Residual Networks (3D ResNets) pre-trained on the Kinetics dataset for measuring the quality of stereoscopic videos and propose a no-reference SVQA method. Specifically, our aim is twofold: Firstly, we answer the question: can we use 3D CNNs as a quality-aware feature extractor from stereoscopic videos or not. Secondly, we explore which ResNet architecture is more appropriate for SVQA. Experimental results on two publicly available SVQA datasets of LFOVIAS3DPh2 and NAMA3DS1-COSPAD1 show the effectiveness of the proposed transfer learning-based method for SVQA that provides the RMSE of 0.332 in LFOVIAS3DPh2 dataset. Also, the results show that deeper 3D ResNet models extract more efficient quality-aware features.Publication Metadata only DeepPyNet: A Deep Feature Pyramid Network for Optical Flow Estimation(IEEE, 2021) Jeny, Afsana Ahsan; Islam, Md Baharul; Aydin, Tarkan; Cree, MJ; Bahcesehir UniversityRecent advances in optical flow prediction have been made possible by using feature pyramids and iterative refining. Though downsampling in feature pyramids may cause foreground items to merge with the background, the iterative processing could be incorrect in optical flow experiments. Particularly the outcomes of the movement of narrow and tiny objects can be more invisible in the flow scene. We introduce a novel method called DeepPyNet for optical flow estimation that includes feature extractor, multi-channel cost volume, and flow decoder. In this method, we propose a deep recurrent feature pyramid-based network for the end-to-end optical flow estimation. The feature extraction from each pixel of the feature map keeps essential information without modifying the feature receptive field. Then, a multi-scale 4 Dc orrelation volume is built from the visual similarity of each pair of pixels. Finally, we utilize the multi-scale correlation volumes to continuously update the flow field through an iterative recurrent method. Experimental results demonstrate that DeepPyNet significantly eliminates flow errors and provides state-of-the-art performance in various datasets. Moreover, DeepPyNet is less complex and uses only 6.1M parameters 81% and 35% smaller than the popular FlowNet and PWC-Net+, respectively.Publication Metadata only PDS-Net: A novel point and depth-wise separable convolution for real-time object detection(SPRINGER, 2022) Junayed, Masum Shah; Islam, Md Baharul; Imani, Hassan; Aydin, Tarkan; Bahcesehir UniversityNumerous recent object detectors and classifiers have shown acceptable performance in recent years by using convolutional neural networks and other efficient architectures. However, most of them continue to encounter difficulties like overfitting, increased computational costs, and low efficiency and performance in real-time scenarios. This paper proposes a new lightweight model for detecting and classifying objects in images. This model presents a backbone for extracting in-depth features and a spatial feature pyramid network (SFPN) for accurately detecting and categorizing objects. The proposed backbone uses point-wise separable (PWS) and depth-wise separable convolutions, which are more efficient than standard convolution. The PWS convolution utilizes a residual shortcut link to reduce computation time. We also propose a SFPN that comprises concatenation, transformer encoder-decoder, and feature fusion modules, which enables the simultaneous processing of multi-scale features, the extraction of low-level characteristics, and the creation of a pyramid of features to increase the effectiveness of the proposed model. The proposed model outperforms all of the existing backbones for object detection and classification in three publicly accessible datasets: PASCAL VOC 2007, PASCAL VOC 2012, and MS-COCO. Our extensive experiments show that the proposed model outperforms state-of-the-art detectors, with mAP improvements of 2.4% and 2.5% on VOC 2007, 3.0% and 2.6% on VOC 2012, and 2.5% and 3.6% on MS-COCO in the small and large sizes of the images, respectively. In the MS-COCO dataset, our model achieves FPS of 39.4 and 33.1 in a single GPU for the small (320 x 320) and large (512 x 512) sizes of the images, respectively, which shows that our method can run in real-time.Publication Metadata only DeepPyNet: A Deep Feature Pyramid Network for Optical Flow Estimation(IEEE Computer Society, 2021) Jeny, Afsana Ahsan; Islam, Md Baharul; Aydin, Tarkan; Cree, M.J.; Jeny, Afsana Ahsan, Department of Computer Engineering, Bahçeşehir Üniversitesi, Istanbul, Turkey; Islam, Md Baharul, Department of Computer Engineering, Bahçeşehir Üniversitesi, Istanbul, Turkey; Aydin, Tarkan, Department of Computer Engineering, Bahçeşehir Üniversitesi, Istanbul, TurkeyRecent advances in optical flow prediction have been made possible by using feature pyramids and iterative refining. Though downsampling in feature pyramids may cause foreground items to merge with the background, the iterative processing could be incorrect in optical flow experiments. Particularly the outcomes of the movement of narrow and tiny objects can be more invisible in the flow scene. We introduce a novel method called DeepPyNet for optical flow estimation that includes feature extractor, multi-channel cost volume, and flow decoder. In this method, we propose a deep recurrent feature pyramid-based network for the end-to-end optical flow estimation. The feature extraction from each pixel of the feature map keeps essential information without modifying the feature receptive field. Then, a multi-scale 4D correlation volume is built from the visual similarity of each pair of pixels. Finally, we utilize the multi-scale correlation volumes to continuously update the flow field through an iterative recurrent method. Experimental results demonstrate that DeepPyNet significantly eliminates flow errors and provides state-of-the-art performance in various datasets. Moreover, DeepPyNet is less complex and uses only 6.1M parameters 81% and 35% smaller than the popular FlowNet and PWC-Net+, respectively. © 2022 Elsevier B.V., All rights reserved.Publication Metadata only Real-Time YOLO-based Heterogeneous front vehicles detection(Institute of Electrical and Electronics Engineers Inc., 2021) Junayed, Masum Shah; Islam, Md Baharul; Sadeghzadeh, Arezoo; Aydin, Tarkan; Kilimci, Z.H.; Yildirim, T.; Piuri, V.; Czarnowski, I.; Camacho, D.; Manolopoulos, Y.; Solak, S.; Junayed, Masum Shah, Department of Computer Engineering, Bahçeşehir Üniversitesi, Istanbul, Turkey, Daffodil International University, Dhaka, Bangladesh; Islam, Md Baharul, Bahçeşehir Üniversitesi, Istanbul, Turkey, College of Data Science and Engineering, American University of Malta, Cospicua, Malta; Sadeghzadeh, Arezoo, Department of Computer Engineering, Bahçeşehir Üniversitesi, Istanbul, Turkey; Aydin, Tarkan, Department of Computer Engineering, Bahçeşehir Üniversitesi, Istanbul, TurkeyThe perception of the complex road environment is a critical factor in autonomous driving, which has become the research focus in intelligent vehicles. In this paper, a real-time front vehicle detection system is proposed to ensure safe driving in a complex environment, particularly in congested megacities. This system is based on the YOLO model, which effectively detects and classifies various vehicles from both images and videos. It improves detection accuracy by modifying a feature extraction-based backbone. To the authors' best knowledge, this is the first time that vehicle detection is implemented on the recently published DhakaAI dataset. Compared to the other available datasets for object detection, such as KITTI, the DhakaAI dataset has a complex environment with numerous vehicles (21 different types). Experimental results demonstrate that the proposed system outperforms the state-of-the-art object detectors. In this method, the mAP (mean average precision) and the FPS (frame per second) is increased by 2.97% and 1.47, 4.64% and 5.57, 4.75% and 3.02, compared to the RetinaNet, SSD, and Faster RCNN on this dataset, respectively. © 2021 Elsevier B.V., All rights reserved.Publication Metadata only HiMODE: A Hybrid Monocular Omnidirectional Depth Estimation Model(IEEE Computer Society, 2022) Junayed, Masum Shah; Sadeghzadeh, Arezoo; Islam, Md Baharul; Lai-Kuan, Wong; Aydin, Tarkan; Junayed, Masum Shah, Bahçeşehir Üniversitesi, Istanbul, Turkey; Sadeghzadeh, Arezoo, Bahçeşehir Üniversitesi, Istanbul, Turkey; Islam, Md Baharul, Bahçeşehir Üniversitesi, Istanbul, Turkey, American University of Malta, Cospicua, Malta; Lai-Kuan, Wong, Multimedia University, Cyberjaya, Malaysia; Aydin, Tarkan, Bahçeşehir Üniversitesi, Istanbul, TurkeyMonocular omnidirectional depth estimation is receiving considerable research attention due to its broad applications for sensing 360° surroundings. Existing approaches in this field suffer from limitations in recovering small object details and data lost during the ground-truth depth map acquisition. In this paper, a novel monocular omnidirectional depth estimation model, namely HiMODE is proposed based on a hybrid CNN+Transformer (encoder-decoder) architecture whose modules are efficiently designed to mitigate distortion and computational cost, without performance degradation. Firstly, we design a feature pyramid network based on the HNet block to extract high-resolution features near the edges. The performance is further improved, benefiting from a self and cross attention layer and spatial/temporal patches in the Transformer encoder and decoder, respectively. Besides, a spatial residual block is employed to reduce the number of parameters. By jointly passing the deep features extracted from an input image at each backbone block, along with the raw depth maps predicted by the transformer encoder-decoder, through a context adjustment layer, our model can produce resulting depth maps with better visual quality than the ground-truth. Comprehensive ablation studies demonstrate the significance of each individual module. Extensive experiments conducted on three datasets, Stanford3D, Matterport3D, and SunCG, demonstrate that HiMODE can achieve state-of-the-art performance for 360° monocular depth estimation. Complete project code and supplementary materials are available at https://github.com/himode5008/HiMODE. © 2025 Elsevier B.V., All rights reserved.