Site Maintenance will take place from 4:00 PM on 2024-04-29 to 9:00 AM on 2024-05-01.
Please do not make any content change during this time, otherwise all the changes will be lost.

Researcher

Mr Mo Hossny

My Expertise

Applied AI in human factors

Biography

Mohammed Hossny received his bachelor degree in computer science from Cairo University. He earned his masters degree in computer science in collaboration with the IBM Centre of Advance Studies (CAS). Later, he joined the Institute for Intelligent Systems and Research Innovation (IISRI) at Deakin University and developed an algebraic framework for multimodal image fusion as part of his PhD degree. He is currently a senior lecturer at the...view more

Mohammed Hossny received his bachelor degree in computer science from Cairo University. He earned his masters degree in computer science in collaboration with the IBM Centre of Advance Studies (CAS). Later, he joined the Institute for Intelligent Systems and Research Innovation (IISRI) at Deakin University and developed an algebraic framework for multimodal image fusion as part of his PhD degree. He is currently a senior lecturer at the University of New South Wales (UNSW). His research is focused around human performance. He is approaching this from marker less motion capture and biomechanics perspectives.


My Research Supervision


Supervision keywords


Areas of supervision

Machine Learning

Artificial Intelligence

View less

Location

UNSW Canberra

Follow

My LinkedIn
My Google Scholar

ORCID as entered in ROS

orcid_icon https://orcid.org/0000-0002-1593-6296

Videos

The VoxelScape #LiDAR #dataset is released (links below). The dataset includes 100K samples of simulated 64-channels #velodynelidar #pointclouds, labels, intensity and bounding boxes. Intensity (from real data) is at the sub-mesh level and is driven by #PBR #material #shaders. Labels are at object and mesh levels. The dataset features a wide variety of vehicles, pedestrians, vegetation, roadwork setups, buildings and procedurally generated #urban landscapes. This is a great collaboration with Khaled Saleh, Ahmed Abobakr, Attia Mohammed and Julie Iskander.
Check out how our refined #rl continuous control via parameterised Tanh performs on the #OpenAI #GYM problems. Paper: https://bit.ly/3oCJIQH

A great collaboration with Julie Iskander, Attia Mohammed and Khaled Saleh. #AI #ML #ArtificialIntelligence #MachineLearning #DeepLearning #MLP#researchpaper #Control #DRL #reinforcementlearning #research #RL #signal
Check out how our #AI #rl agent handles different types of noise while maintaining a standing posture. The agent was trained on levelled ground with no noise and was able to accommodate different noise sources. A great collaboration with Julie Iskander.
Value misalignment strikes back [1]. This AI agent learned to walk backwards to prolong not falling. A great collaboration with Julie Iskander.

Links
[1] https://bit.ly/3lktTfr
Check out how our #AI #rl agent handles different types of noise while maintaining a standing posture. The agent was trained on levelled ground with no noise and was able to accommodate different noise sources. A great collaboration with Julie Iskander.
VoxelScape Dataset
AI learns Refined Control
An AI agent achieves postural balnace
Value misalignment strikes back
An intoxicated AI agent learns to stand with noisy states and actions

UNSW Sydney NSW 2052 Australia Telephone +61 2 93851000
Authorised by Deputy Vice-Chancellor (Research)
UNSW CRICOS Provider Code: 00098G ABN: 57 195 873 179

Join the conversation