A5040389900- Image Enhancement Techniques
- Power Line Inspection Robots
- Advanced Image Processing Techniques
- Robotics and Sensor-Based Localization
- Advanced Vision and Imaging
- Mechanical and Optical Resonators
- Force Microscopy Techniques and Applications
- Multimodal Machine Learning Applications
- Advanced MEMS and NEMS Technologies
- Topic Modeling
- Face recognition and analysis
- Face and Expression Recognition
- Image and Video Stabilization
- Image and Object Detection Techniques
- Emotion and Mood Recognition
- Robotic Path Planning Algorithms
- Sentiment Analysis and Opinion Mining
- Soft Robotics and Applications
- Speech and Audio Processing
American University of Sharjah
2019-2023
Conventionally, aerial manipulators, when used for inspection, use drone rotors to stabilize the center of gravity (CoG) shifts, which highly affects its performance. This paper discusses development a self-balancing lightweight cable manipulator that can be construction inspection purposes. The design is based on 3D-printed, three degrees freedom (DoF), planar mounted an extended platform below it as counter-balance mechanism. actuators control links through system, allowing them at system...
Computing optical flow is a fundamental problem in computer vision. However, deep learning-based techniques do not perform well for non-rigid movements such as those found faces, primarily due to lack of the training data representing fine facial motion. We hypothesize that learning on face motion will improve quality predicted faces. This work aims to: (1) exploring self-supervised generate ground truth images; (2) computing baseline results effects using train Convolutional Neural Networks...
In this research, a novel design of lightweight aerial manipulator system is proposed for solar panel cleaning with active (CoG) compensation mechanism. Recently, separate arrays or units are commonly installed on residential, commercial rooftops roads, making it inconvenient land robots to perform the tasks. The light weight gravity mechanism intended be attached beneath drone increase its stability during operation. workspace given analyzed under CoG shift constraints. kinematics and...
In this work we propose additions to the COSMOS and on Steroids pipelines for detection of Cheapfakes Task 1 ACM Grand Challenge Detecting Cheapfakes. We compute sentiment features, namely polarity subjectivity, using news image captions. Multiple logistic regression results show that these features are significant in prediction outcome. then combine with four image-text obtained aforementioned previous works train an MLP. This classifies sets inputs into being out-of-context (OOC) or...
Computing optical flow is a fundamental problem in computer vision. However, deep learning-based techniques do not perform well for non-rigid movements such as those found faces, primarily due to lack of the training data representing fine facial motion. We hypothesize that learning on face motion will improve quality predicted faces. The aim this work threefold: (1) exploring self-supervised generate ground truth images; (2) computing baseline results effects using train Convolutional...