Browsing by Author "Ammar, Adel"
Now showing 1 - 10 of 13
Results Per Page
Sort Options
- Activity Monitoring of Islamic Prayer (Salat) Postures using Deep LearningPublication . Koubaa, Anis; Ammar, Adel; Benjdira, Bilel; Al Hadid, Abdullatif; Kawaf, Belal; Al Yahri, Saleh Ali; Babiker, Abdelrahman; Assaf, Koutaiba; Ba Ras, MohannadIn the Muslim community, the prayer (i.e. Salat) is the second pillar of Islam, and it is the most essential and fundamental worshiping activity that believers have to perform five times a day. From a gestures' perspective, there are predefined human postures that must be performed in a precise manner. However, for several people, these postures are not correctly performed, due to being new to Salat or even having learned prayers in an incorrect manner. Furthermore, the time spent in each posture has to be balanced. To address these issues, we propose to develop an artificial intelligence assistive framework that guides worshippers to evaluate the correctness of the postures of their prayers. This paper represents the first step to achieve this objective and addresses the problem of the recognition of the basic gestures of Islamic prayer using Convolutional Neural Networks (CNN). The contribution of this paper lies in building a dataset for the basic Salat positions, and train a YOLOv3 neural network for the recognition of the gestures. Experimental results demonstrate that the mean average precision attains 85% for a training dataset of 764 images of the different postures. To the best of our knowledge, this is the first work that addresses human activity recognition of Salat using deep learning.
- AI-based Pilgrim Detection using Convolutional Neural NetworksPublication . Jabra, Marwa Ben; Ammar, Adel; Koubaa, Anis; Cheikhrouhou, Omar; Hamam, HabibPilgrimage represents the most important Islamic religious gathering in the world where millions of pilgrims visit the holy places of Makkah and Madinah to perform their rituals. The safety and security of pilgrims is the highest priority for the authorities. In Makkah, 5000 cameras are spread around the holy mosques for monitoring pilgrims, but it is almost impossible to track all events by humans considering the huge number of images collected every second. To address this issue, we propose to use an artificial intelligence technique based on deep learning and convolutional neural networks to detect and identify Pilgrims and their features. For this purpose, we built a comprehensive dataset for the detection of pilgrims and their genders. Then, we develop two convolutional neural networks based on YOLOv3 and Faster-RCNN for the detection of Pilgrims. Experiment results show that Faster RCNN with Inception v2 feature extractor provides the best mean average precision over all classes (51%). A video demonstration that illustrates a real-time pilgrim detection using our proposed model is available at [1].
- Car Detection using Unmanned Aerial Vehicles: Comparison between Faster R-CNN and YOLOv3Publication . Benjdira, Bilel; Khursheed, Taha; Koubaa, Anis; Ammar, Adel; Ouni, KaisUnmanned Aerial Vehicles are increasingly being used in surveillance and traffic monitoring thanks to their high mobility and ability to cover areas at different altitudes and locations. One of the major challenges is to use aerial images to accurately detect cars and count-them in real-time for traffic monitoring purposes. Several deep learning techniques were recently proposed based on convolution neural network (CNN) for real-time classification and recognition in computer vision. However, their performance depends on the scenarios where they are used. In this paper, we investigate the performance of two state-of-the art CNN algorithms, namely Faster R-CNN and YOLOv3, in the context of car detection from aerial images. We trained and tested these two models on a large car dataset taken from UAVs. We demonstrated in this paper that YOLOv3 outperforms Faster R-CNN in sensitivity and processing time, although they are comparable in the precision metric.
- Data-Efficient Domain Adaptation for Semantic Segmentation of Aerial Imagery Using Generative Adversarial NetworksPublication . Ben Jdira, Bilel; Ammar, Adel; Koubaa, Anis; Ouni, KaïsDespite the significant advances noted in semantic segmentation of aerial imagery, a considerable limitation is blocking its adoption in real cases. If we test a segmentation model on a new area that is not included in its initial training set, accuracy will decrease remarkably. This is caused by the domain shift between the new targeted domain and the source domain used to train the model. In this paper, we addressed this challenge and proposed a new algorithm that uses Generative Adversarial Networks (GAN) architecture to minimize the domain shift and increase the ability of the model to work on new targeted domains. The proposed GAN architecture contains two GAN networks. The first GAN network converts the chosen image from the target domain into a semantic label. The second GAN network converts this generated semantic label into an image that belongs to the source domain but conserves the semantic map of the target image. This resulting image will be used by the semantic segmentation model to generate a better semantic label of the first chosen image. Our algorithm is tested on the ISPRS semantic segmentation dataset and improved the global accuracy by a margin up to 24% when passing from Potsdam domain to Vaihingen domain. This margin can be increased by addition of other labeled data from the target domain. To minimize the cost of supervision in the translation process, we proposed a methodology to use these labeled data efficiently.
- DeepBrain: Experimental Evaluation of Cloud-Based Computation Offloading and Edge Computing in the Internet-of-Drones for Deep Learning ApplicationsPublication . Koubaa, Anis; Ammar, Adel; Alahda, Mahmoud; Kanhouc, Anas; Azar, Ahmad TaherUnmanned Aerial Vehicles (UAVs) have been very effective in collecting aerial images data for various Internet-of-Things (IoT)/smart cities applications such as search and rescue, surveillance, vehicle detection, counting, intelligent transportation systems, to name a few. However, the real-time processing of collected data on edge in the context of the Internet-of-Drones remains an open challenge because UAVs have limited energy capabilities, while computer vision techniquesconsume excessive energy and require abundant resources. This fact is even more critical when deep learning algorithms, such as convolutional neural networks (CNNs), are used for classification and detection. In this paper, we first propose a system architecture of computation offloading for Internet-connected drones. Then, we conduct a comprehensive experimental study to evaluate the performance in terms of energy, bandwidth, and delay of the cloud computation offloading approach versus the edge computing approach of deep learning applications in the context of UAVs. In particular, we investigate the tradeoff between the communication cost and the computation of the two candidate approaches experimentally. The main results demonstrate that the computation offloading approach allows us to provide much higher throughput (i.e., frames per second) as compared to the edge computing approach, despite the larger communication delays.
- Design and performance analysis of global path planning techniques for autonomous mobile robots in grid environmentsPublication . Chaari, Imen; Koubâa, Anis; Bennaceur, Hachemi; Ammar, Adel; Alajlan, Maram; Youssef, HabibThis article presents the results of the 2-year iroboapp research project that aims at devising path planning algorithms for large grid maps with much faster execution times while tolerating very small slacks with respect to the optimal path. We investigated both exact and heuristic methods. We contributed with the design, analysis, evaluation, implementation and experimentation of several algorithms for grid map path planning for both exact and heuristic methods. We also designed an innovative algorithm called relaxed A-star that has linear complexity with relaxed constraints, which provides near-optimal solutions with an extremely reduced execution time as compared to A-star. We evaluated the performance of the different algorithms and concluded that relaxed A-star is the best path planner as it provides a good trade-off among all the metrics, but we noticed that heuristic methods have good features that can be exploited to improve the solution of the relaxed exact method. This led us to design new hybrid algorithms that combine our relaxed A-star with heuristic methods which improve the solution quality of relaxed A-star at the cost of slightly higher execution time, while remaining much faster than A* for large-scale problems. Finally, we demonstrate how to integrate the relaxed A-star algorithm in the robot operating system as a global path planner and show that it outperforms its default path planner with an execution time 38% faster on average.
- Global robot Path Planning using GA for Large Grid Maps: Modelling, Performance and ExperimentationPublication . Alajlan, Maram; Chaari, Imen; Koubâa, Anis; Bennaceur, Hachemi; Ammar, Adel; Youssef, HabibIn this paper, the efficiency of genetic algorithm (GA) approach to address the problem of global path planning for mobile robots in large-scale grid environments is revisited and assessed. First, an efficient GA path planner to find an (or near) optimal path in a grid map is proposed. In particular, large maps instances are considered in this work, as small maps are easy to address by typical linear-time exact algorithms, in contrast to large maps, which require more intensive computations. The operators of the GA planner were carefully designed for optimizing the search process. Also, extensive simulations to evaluate the GA planner are conducted, and its performance is compared to that of the A algorithm considered as benchmarking reference. We found out that the GA planner can find optimal solutions in the same way as A in large grid maps in most cases, but A is faster than the GA. This is because GA is not a constructive path planner and heavily relies on initial population to explore the space of solutions in contrast to A that builds its solution progressively towards the target. It was concluded that, although GA can provide an alternative to A technique, it is likely that they are not efficient enough to beat exact methods such as A when used with a consistent heuristic. The GA planner is integrated in the global path planning modules of the Robot Operating System (ROS), its feasibility is demonstrated, and its performance is compared against the default ROS planner.
- A Hybrid Deep Learning Model for UAVs Detection in Day and Night Dual VisionsPublication . Noor, Alam; Li, Kai; Ammar, Adel; Koubâa, Anis; Benjdira, Bilel; Tovar, EduardoUnmanned Aerial Vehicle (UAV) detection for public safety protection is becoming a critical issue in non-fly zones. There are plenty of attempts of the UAV detection using single stream (day or night vision). In this paper, we propose a new hybrid deep learning model to detect the UAVs in day and night visions with a high detection precision and accurate bounding box localization. The proposed hybrid deep learning model is developed with cosine annealing and rethinking transformation to improve the detection precision and accelerate the training convergence. To validate the hybrid deep learning model, real-world experiments are conducted outdoor in daytime and nighttime, where a surveillance video camera on the ground is set up for capturing the UAV. In addition, the UAV-Catch open database is adopted for offline training of the proposed hybrid model, which enriches training datasets and improves the detection precision. The experimental results show that the proposed hybrid deep learning model achieves 65% in terms of the mean average detection precision given the input videos in day and night visions.
- Move and Improve: a Market-Based Mechanism for the Multiple Depot Multiple Travelling Salesmen ProblemPublication . Koubâa, Anis; Cheikhrouhou, Omar; Bennaceur, Hachemi; Sriti, Mohamed-Foued; Javed, Yasir; Ammar, AdelConsider the problem of having a team of cooperative and autonomous robots to repeatedly visit a set of target locations and return back to their initial locations. This problem is known as multi-robot patrolling and can be cast to the multiple depot multiple traveling salesman problem (MD-MTSP), which applies to several mobile robots applications. As an NP-Hard problem, centralized approaches using meta-heuristic search are typically used to solve it, but such approaches are computation-intensive and cannot effectively deal with the dynamic nature of the system. This paper provides a distributed solution based on a market-based approach, called Move-and-Improve. It involves the cooperation of the robots to incrementally allocate targets and remove possible overlap. The concept is simple: in each step, a robot moves and attempts to improve its solution while communicating with its neighbors. Our approach consists of four main phases: (1) initial target allocation, (2) tour construction, (3) negotiation of conflicting targets, (4) solution improvement. To validate the efficiency of the Move-and-Improve distributed algorithm, we first conducted extensive simulations using Webots and evaluated its performance in terms of total traveled distance, maximum tour length, and ratio of overlapped targets, under different settings. We also demonstrated through MATLAB simulations the benefits of using our decentralized approach as compared to a centralized Genetic Algorithm approach to solve the MD-MTSP problem. Finally, we implemented Move-and-Improve using ROS and deployed it on real robots.
- Move and Improve: a Market-Based Mechanism for the Multiple Depot Multiple Travelling Salesmen ProblemPublication . Koubâa, Anis; Cheikhrouhou, Omar; Bennaceur, Hachemi; Sriti, Mohamed-Foued; Javed, Yasir; Ammar, AdelConsider the problem of having a team of cooperative and autonomous robots to repeatedly visit a set of target locations and return back to their initial locations. This problem is known as multi-robot patrolling and can be cast to the multiple depot multiple traveling salesman problem (MD-MTSP), which applies to several mobile robots applications. As an NP-Hard problem, centralized approaches using meta-heuristic search are typically used to solve it, but such approaches are computation-intensive and cannot effectively deal with the dynamic nature of the system. This paper provides a distributed solution based on a market-based approach, called Move-and-Improve. It involves the cooperation of the robots to incrementally allocate targets and remove possible overlap. The concept is simple: in each step, a robot moves and attempts to improve its solution while communicating with its neighbors. Our approach consists of four main phases: (1) initial target allocation, (2) tour construction, (3) negotiation of conflicting targets, (4) solution improvement. To validate the efficiency of the Move-and-Improve distributed algorithm, we first conducted extensive simulations using Webots and evaluated its performance in terms of total traveled distance, maximum tour length, and ratio of overlapped targets, under different settings. We also demonstrated through MATLAB simulations the benefits of using our decentralized approach as compared to a centralized Genetic Algorithm approach to solve the MD-MTSP problem. Finally, we implemented Move-and-Improve using ROS and deployed it on real robots.