Mohamed Hasan

University of Leeds · School of Computing

Human sensorimotor decision making has a tendency to get ‘stuck in a rut’, being biased towards selecting a previously implemented action structure (hysteresis). Existing explanations propose this is the consequence of an agent efficiently modifying an existing plan, rather than creating a new plan from scratch. Instead, we propose that hysteresis...

Autonomous vehicles use a variety of sensors and machine-learned models to predict the behavior of surrounding road users. Most of the machine-learned models in the literature focus on quantitative error metrics like the root mean square error (RMSE) to learn and report their models’ capabilities. This focus on quantitative error metrics tends to i...

Autonomous vehicles use a variety of sensors and machine-learned models to predict the behavior of surrounding road users. Most of the machine-learned models in the literature focus on quantitative error metrics like the root mean square error (RMSE) to learn and report their models' capabilities. This focus on quantitative error metrics tends to i...

Autonomous vehicles should be able to predict the future states of its environment and respond appropriately. Specifically, predicting the behavior of surrounding human drivers is vital for such platforms to share the same road with humans. Behavior of each of the surrounding vehicles is governed by the motion of its neighbor vehicles. This paper f...

Predicting future behavior of the surrounding vehicles is crucial for self-driving platforms to safely navigate through other traffic. This is critical when making decisions like crossing an unsignalized intersection. We address the problem of vehicle motion prediction in a challenging roundabout environment by learning from human driver data. We e...

There is quickly growing literature on machine-learned models that predict human driving trajectories in road traffic. These models focus their learning on low-dimensional error metrics, for example average distance between model-generated and observed trajectories. Such metrics permit relative comparison of models, but do not provide clearly inter...

Human sensorimotor decision-making has a tendency to get ‘stuck in a rut’, being biased towards selecting a previously implemented action structure (‘hysteresis’). Existing explanations cannot provide a principled account of when hysteresis will occur. We propose that hysteresis is an emergent property of a dynamical system learning from the conseq...

The objective of this project is learning high-level manipulation planning skills from humans and transfer these skills to robot planners. We used virtual reality to generate data from human participants whilst they reached for objects on a cluttered table top. From this, we devised a qualitative representation of the task space to abstract human d...

Humans, in comparison to robots, are remarkably adept at reaching for objects in cluttered environments. The best existing robot planners are based on random sampling of configuration space -- which becomes excessively high-dimensional with large number of objects. Consequently, most planners often fail to efficiently find object manipulation plans...

Humans, in comparison to robots, are remarkably adept at reaching for objects in cluttered environments. The best existing robot planners are based on random sampling in configuration space- which becomes excessively highdimensional with a large number of objects. Consequently, most of these planners suffer from limited object manipulation. We addr...

This paper introduces the Gait Gate as the first online walk-through access control system based on multimodal biometric person verification. Face, gait and height modalities are simultaneously captured by a single RGB-D sensor and fused at the matching-score level. To achieve the real-time requirements, mutual subspace method has been used for the...

Anthropometric biometrics are the most suitable features for long-term people reidentification. However, feature selection is still a major problem in the anthropometric biometrics literature. In this paper, we aim at improving people reidentification by enhancing feature selection. Based on a statistical analysis of body measurements on a large-sc...

This paper proposes a new technique for rectangle detection in images based on hierarchical feature complexity. The algorithm follows a bottom-up/top-down approach: in the bottom-up phase, contour curves are extracted and its edges are fit to straight lines. Long contours may grow away from the object boundary and they may not complete a loop due t...

this paper verifies a recently published method of monocular depth computing in the context of visual SLAM. The closed form depth solution was exploited in the measurement model of a monocular EKF visual SLAM algorithm. SIFT interest points are tracked during camera motion and a suitable feature initialization is presented. The visual SLAM system i...

Monocular depth has been found using estimation, closed-form solution and learning techniques. Estimation and closed-form solution compute the depth from motion, while learning techniques calculate the depth using a single image with a depth map as a supervisor. This paper presents a new closed form solution for monocular depth from motion. The pro...

The majority of visual SLAM techniques utilize interest points as landmarks. Therefore, they suffer from two main problems; scalability and data association reliability. Recently, there has been increasing interest in using higher level object description to reduce the number of tracked features and improve the data association among frames. In thi...

We consider the problem of controlling the vertical motion of a nonlinear model of a helicopter during landing, while stabilizing the lateral and longitudinal position and maintaining a constant attitude. The dynamics of the helicopter main and tail rotors are often neglected to simplify the control design, but this reduction has the disadvantage o...