Various research video demos with links to available open access manuscripts, open source software and datasets.

Robust Cell Detection and Classification in Whole Slide Images Using Neural Graph Refinement

Classifying nuclei communities in histology images is vital for early cancer treatment, but it remains challenging due to the similar structure of nuclei communities. To address this problem, we proposed an iterative neural graph improvement and broadcasting approach. A fully connected graph is constructed with nuclei as nodes starting with a baseline classification. Node and edge features are updated and exchanged along a Hamiltonian path, removing weak connections. This process filters communities by disconnecting weakly connected nodes and iterates until stability is reached. Loose nodes from this refining stage are then assigned to their closest community clusters.

More in: Hassan, T., Li, Z., Javed, S., Dias, J. and Werghi, N., 2023. Neural Graph Refinement for Robust Recognition of Nuclei Communities in Histopathological Landscape. IEEE Transactions on Image Processing. [pdf]

Robot-Person Tracking in Uniform Appearance Scenarios: A New Dataset and Challenges

Person-tracking robots have many applications, including surveillance and autonomous driving. However, Despite the abundance of uniform appearance in many contexts and the challenges they exhibit, there is a lack of video datasets dedicated to benchmarking tracking algorithms in such contexts. In this work, we proposed a new high-quality RGB-D benchmark called PTUA for robot–person tracking in uniform appearance scenarios. PTUA is recorded using an RGB-D sensor on top of a moving robot and consists of 45 sequences containing more than 85 K frames. Each frame is manually annotated with a bounding box and attributes, making PTUA the largest and the most challenging person tracking RGB-D dataset.

More in: Zhang, X., Ghimire, A., Javed, S., Dias, J. and Werghi, N. “Robot-Person Tracking in Uniform Appearance Scenarios: A New Dataset and Challenges” IEEE Transactions on Human-Machine Systems, 2023.

Drone-Person Tracking in a Uniform Appearance Crowd: A New Dataset

Drone-person tracking in uniform appearance crowds poses unique challenges due to the difficulty in distinguishing individuals with similar Replace this image with the video at this link: attire over multi-scale variations. To address this issue and facilitate the development of effective tracking algorithms, we present a novel dataset named D-PTUAC (Drone-Person Tracking in Uniform Appearance Crowd). The dataset comprises 138 sequences comprising over 121K frames, each manually annotated with bounding boxes and attributes. During dataset creation, we carefully consider 18 challenging attributes encompassing a wide range of viewpoints and scene complexities. These attributes are annotated to facilitate the analysis of performance based on specific attributes.

More in: M Alansari, OA Hay, S Javed, A Shoufan, Y Zweiri, N Werghi, 2024. Drone-Person Tracking in Uniform
Appearance Crowd: A New Dataset. Scientific Data, Nature Springer. [pdf]

Detecting threat items: An anomaly approach

In this study, we developed a luggage screening system that categorizes baggage as either normal (containing a threat item) or abnormal (free of threat items). We introduced a single- stage detection approach, where an encoder- decoder model is trained to reconstruct the content of normal baggage. The model then identifies abnormal regions by examining disparities between the original and reconstructed scans.

More in: Hassan, T., Akcay, S., Bennamoun, M., Khan, S. and Werghi, N., 2021. Unsupervised anomaly instance segmentation for
baggage threat recognition. Journal of Ambient Intelligence and Humanized Computing, pp.1-12.  [pdf]

Detecting, localizing and identifying threat items

In this work, we designed an object detector model capable of pinpointing and identifying threat items within baggage. Unlike traditional detectors, our model utilizes a multi-resolution contour map representation, enabling it to overcome the challenge of limited texture in X-ray scans.

More in: Hassan, T., Akcay, S., Hassan, B., Bennamoun, M., Khan, S., Dias, J. and Werghi, N.. Cascaded structure tensor
for robust baggage threat detection. Neural Computing and Applications, 35(15), 2023. [pdf]

Detecting, Localizing and Identifying Threat Items: an Incremental Semantic Segmentation approach

In this research, we pinpoint and identify threat items by isolating their specific regions in the baggage scan. The proposed model stands out in two key aspects: 1) its ability to handle multiple instances of threats, even in complex configurations (such as a set of overlapping pistols); and 2) its capability to be deployed incrementally, allowing the model to accommodate new threat instances without the necessity of retraining from scratch.

More in: Hassan, T., Akcay, S., Bennamoun, M., Khan, S. and Werghi, N., 2021. A novel incremental learning driven instance segmentation framework to recognize highly cluttered instances of the contraband items. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 52(11), pp.6937-6951.  [pdf]

Autonomous Inspection of Flare Stacks using Unmanned Aerial Vehicles

Flare stack systems are a crucial component in the operation of oil refineries and petrochemical plants as they safely release the excess gas generated during the plant’s operation. Performance and structure inspection of these systems are essential and challenging tasks due to the flare stacks’ Replace this image with the video at this link harsh operating environment. Flare stacks go through various types of faults both in their mechanical structure and combustion operation, including cracks in the structure, abnormal pilot flame operation, and incomplete combustion of the released gas. The occurrence of such faults could lead to leakage of dangerous emissions to the environment and deadly fires and explosions in the petrochemical plant. Autonomous robotic systems in the inspection of such systems is a promising solution for minimizing the involved hazards and costs. In this work, we proposed an autonomous, low-cost unmanned aerial vehicle system (UAVS) for inspecting the flare stack. The proposed UAVS uses the video stream obtained from an on-board camera for analyzing the observed scene, controlling the drone’s movement, and assessing the the flare stack’s operation.

More in: M AlRadi, H Karki, N Werghi, S Javed, J.Dias.
2023. Multi-view Inspection of Flare Stacks Operation Using a Vision-controlled Autonomous UAV.
IECON 2022–48th Annual Conference of the IEEE Industrial Electronics Society. 2023.  [pdf]

Scroll to Top