Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Roshan George has a total of 35 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Impact of Inter-Modality Misalignment on Multispectral Object Detection

    by Muhammad Usama, Imad Ali Shah, Roshan George, Enda Ward, Fiachra Collins, William O’GRady, Martin Glavin, Brian Deegan, Edward Jones

    Object Detection significantly benefits from combining complementary spectral modalities; however, its performance suffers greatly from misregistration between sensor streams due to spatial misalignments, viewpoint differences between sensors, and inconsistent annotations, a common feature of multi-modal datasets used in object detection system development. In this work, we address the critical challenge of modality misregistration. We propose a comprehensive alignment framework based on object-aware local homography estimation, anchored on Long-Wave Infrared Radiation (LWIR) annotations, to generate a spatially aligned training dataset for improved model development. For our experiments, we utilise the Multispectral Object Detection (MOD) dataset, which includes RGB, Short-Wave Infrared Radiation (SWIR), Mid-Wave Infrared Radiation (MWIR), and Long-Wave Infrared Radiation (LWIR) imagery. The proposed alignment pipeline ensures a consistent field of view, resolution normalization, and accurate cross-modality annotation projection. To quantify the impact of alignment, we benchmark four state-of-the-art fusion architectures (DEYOLO, TarDAL, CMTFusion, and U2Fusion) on both the original and the aligned versions of the MOD dataset. Experimental results demonstrate significant improvements of 10 to 30 percentage points in mAP@0.5 after training on the aligned MOD data, with enhanced localization accuracy and reduced false detections. We further verified the generalizability of the proposed alignment framework on a different dataset (FLIR ADAS), observing 3.54 percentage points improvement in mAP@0.5 after training on the aligned RGB-LWIR image pairs. U2Fusion emerged as the architecture most tolerant to misregistration in initial experiments, and this architecture is used to conduct a further sensitivity analysis to evaluate model robustness against real-world misalignments, in a controlled way. The results of this sensitivity analysis could be useful for designers of practical systems, where small misalignment errors between image modalities commonly occur.


  • Open Access

    Information Capacity as a Predictor of Perception Performance

    by Diarmaid Geever, Tim Brophy, Dara Molloy, Enda Ward, Roshan George, Norman Koren, Martin Glavin, Edward Jones, Brian Deegan

    The design of automated driving systems is of growing industry and societal interest. Perception is a critical technology for these systems, which allows a vehicle to discern the surrounding environment. Perception systems in automated vehicles frequently use machine vision algorithms; however, the performance of a machine vision algorithm critically depends on the quality of the data provided. Quantifying the ‘quality’ of image data is therefore potentially a useful tool in understanding and predicting the performance of a machine vision system. This study uses the Shannon Information Capacity, a metric based on information theory, to evaluate the impact of image quality on a perception algorithm. In this preliminary study, a set of synthetic objects are arranged to create a novel simulated test chart. The chart contains standard machine vision objects of interest (people and cars) as well as a slanted edge, which is used to calculate image quality metrics. The chart is degraded using varying levels of contrast and blur to simulate different real-world operating conditions. Object detection performance is then evaluated using a range of deep learning-based detection algorithms, with different architectures. The results indicate that Shannon Information Capacity has the potential to predict machine vision performance across multiple model architectures and object types. For example, the results for all the models show that accuracy remains relatively constant above an SIC value of 0.25 b/p. Results indicate that for YOLOv10m SIC has mutual information value with detection accuracy of 1.66 bits while MTF50 has a score of 0.4945 bits. This study is the first to show the correlation between SIC and machine vision performance. While other metrics have been previously shown to have some correlation with machine vision, the correlation shown by SIC is much stronger. The findings presented may be of use to designers of autonomous driving systems and automotive camera manufacturers.


  • Open Access

    Impact of Contrast Threshold on Event-Based Automotive Perception Performance

    by Ethan Delaney, Roshan George, Tim Brophy, Imad Ai Shah, Enda Ward, Edward Jones, Brian Deegan, Martin Glavin

  • Open Access

    TFE-Fusion: Tri-Modality Feature Enhanced Fusion for Robust Object Detection

    by Muhammad Usama, Imad Ali Shah, Roshan George, Enda Ward, Fiachra Collins, William O'GRady, Martin Glavin, Brian Deegan, Edward Jones

  • Open Access

    G-MIND: Galway Multimodal Infrastructure Node Dataset for Intelligent Transportation Systems

    by Dara Molloy, Roshan George, Tim Brophy, Brian Deegan, Darragh Mullins, Enda Ward, Jonathan Horgan, Ciaran Eising, Patrick Denny, Edward Jones, et al.

    Autonomous and semi-autonomous vehicles require accurate perception of their surrounding environment to ensure safe operation, yet onboard sensors frequently encounter occlusion challenges that result in incomplete dynamic environmental maps. Infrastructure-to-vehicle cooperative perception addresses this by deploying infrastructure nodes that monitor scenes and share reliable environmental maps with nearby vehicles via technologies like C-V2X. However, existing infrastructure perspective datasets lack diverse multi-modal data and aerial footage, which are crucial to determine effectively the necessary sensors for safety-critical infrastructure node applications. This paper introduces G-MIND, a multimodal infrastructure node dataset supporting research into sensor suitability for infrastructure-assisted safety-critical applications. G-MIND is the first dataset to incorporate this comprehensive range of sensing modalities for infrastructure-based perception: RGB, FIR, and neuromorphic cameras, LiDARs, RADAR, and aerial drone footage. With 91,500 annotated frames, G-MIND offers a larger scale than existing infrastructure perception datasets such as Ko-PER (10k frames), CoopScenes (40K frames), and DAIR-V2X (71k frames), enabling more comprehensive training and evaluation. The dataset captures day and night scenarios featuring cars, pedestrians, and cyclists across diverse traffic scenarios. Beyond standard perception benchmarking, G-MIND includes specialized collections designed to test perception system boundaries: maximum detection distance scenarios, far and occluded object scenarios, and pedestrian action prediction scenarios that challenge current algorithms. Additionally, this paper analyzes what constitutes effective ITS infrastructure node sensors from a practical perspective, comparing modalities against technical criteria (field of view, spatial resolution, low light performance, adverse weather resilience) and pragmatic criteria (cost, durability). The dataset is available for non-commercial research at https://ieee-dataport.org/documents/galway-multimodal-infrastructure-node-dataset , with the SDK at https://github.com/daramolloy/GMIND-sdk.


  • Open Access

    Hyperspectral Sensors and Autonomous Driving: Technologies, Limitations, and Opportunities

    by Imad Ali Shah, Jiarong Li, Roshan George, Tim Brophy, Enda Ward, Martin Glavin, Edward Jones, Brian Deegan

    Hyperspectral imaging (HSI) is a transformative sensing modality for Advanced Driver Assistance Systems (ADAS) and autonomous driving (AD). By capturing fine spectral resolution across hundreds of bands, HSI enables material-level scene understanding that overcomes critical limitations of traditional RGB imaging in adverse weather and lighting. This paper presents the first comprehensive review of HSI for automotive applications, examining the strengths, limitations, and suitability of current HSI technologies in the context of ADAS/AD. In addition, we analyze 216 commercially available spectral imaging cameras, benchmarking them against key automotive criteria: frame rate, spatial resolution, spectral dimensionality, and compliance with AEC-Q100 temperature standards. Our analysis reveals a significant gap between HSI’s demonstrated research potential and its commercial readiness. Only four cameras meet the defined performance thresholds, and none comply with AEC-Q100 requirements. In addition, the paper reviews recent HSI datasets and applications, including semantic segmentation for road surface classification, pedestrian separability, and adverse weather perception. Our review shows that current HSI datasets are limited in scale, spectral consistency, channel count, and environmental diversity, posing a challenge for perception algorithms development and adequate HSI’s potential validation in ADAS/AD applications. This review paper presents the current state of HSI in automotive contexts and outlines key research directions toward practical integration of spectral imaging in ADAS and autonomous systems.


  • Open Access

    EPCNet: Implementing an ‘Artificial Fovea’ for More Efficient Monitoring Using the Sensor Fusion of an Event-Based and a Frame-Based Camera

    by Orla Sealy Phelan, Dara Molloy, Roshan George, Edward Jones, Martin Glavin, Brian Deegan

    Efficient object detection is crucial to real-time monitoring applications such as autonomous driving or security systems. Modern RGB cameras can produce high-resolution images for accurate object detection. However, increased resolution results in increased network latency and power consumption. To minimise this latency, Convolutional Neural Networks (CNNs) often have a resolution limitation, requiring images to be down-sampled before inference, causing significant information loss. Event-based cameras are neuromorphic vision sensors with high temporal resolution, low power consumption, and high dynamic range, making them preferable to regular RGB cameras in many situations. This project proposes the fusion of an event-based camera with an RGB camera to mitigate the trade-off between temporal resolution and accuracy, while minimising power consumption. The cameras are calibrated to create a multi-modal stereo vision system where pixel coordinates can be projected between the event and RGB camera image planes. This calibration is used to project bounding boxes detected by clustering of events into the RGB image plane, thereby cropping each RGB frame instead of down-sampling to meet the requirements of the CNN. Using the Common Objects in Context (COCO) dataset evaluator, the average precision (AP) for the bicycle class in RGB scenes improved from 21.08 to 57.38. Additionally, AP increased across all classes from 37.93 to 46.89. To reduce system latency, a novel object detection approach is proposed where the event camera acts as a region proposal network, and a classification algorithm is run on the proposed regions. This achieved a 78% improvement over baseline.


  • Open Access

    The Impact of Pose Alignment Errors on a Classical Late Infrastructure-Vehicle Collaboration Framework Using Experimental Data

    by Roshan George, Dara Molloy, Tim Brophy, William O'GRady, Darragh Mullins, Edward Jones, Brian Deegan, Martin Glavin

    Accurate perception and localization of dynamic objects is crucial for autonomous systems to navigate complex environments and avoid collisions. However, onboard sensing struggles with complex environments due to the challenges associated with occlusion, hindering progress in achieving advanced levels of vehicle autonomy in these environments. Cooperative driving automation has emerged as an enabling technology for the development of safer vehicles by using shared perception information, thus enhancing the situational awareness of the ego-vehicle. Alignment of shared perceptual information is a crucial task within this cooperative framework, and any pose errors can compromise the safety of the system. This paper investigates the real-world impact of pose alignment errors on vehicle-infrastructure collaborative driving by identifying, isolating, and analyzing individual error sources within a late collaboration framework. Our proposed method creates a shared global V2X environmental model by fusing LiDAR object representations from individual agents. We examine the sensitivity of this global map to common real-world error sources, including V2X communication delays, GPS positioning uncertainty, and sensor calibration errors. In real-world scenarios, these map alignment errors are non-zero, and their magnitudes depend on the specific use case. By isolating and studying each error type individually, we gain insight into their relative impacts and how they accumulate in real-world collaborative driving systems. We demonstrate the benefits of V2X collaboration and determine the most impactful errors on map misalignment through quantitative analysis, thus allowing the reader to conceptualize the implications and challenges of pose alignment errors in collaborative driving.


  • Open Access

    Infrastructure Assisted Autonomous Driving: Research, Challenges, and Opportunities

    by Roshan George, Joseph Clancy, Tim Brophy, Ganesh Sistu, William O'GRady, Sunil Chandra, Fiachra Collins, Darragh Mullins, Edward Jones, Brian Deegan, et al.

    Despite advancements in perception technology, achieving full autonomy in vehicles remains challenging partly due to limited situational awareness. Even with their sophisticated sensor arrays, autonomous vehicles often struggle to comprehend complex real-world environments due to the challenges associated with occlusion. A possible solution for addressing this limitation lies in the concept of vehicle-to-infrastructure cooperative driving, which enables vehicles to interact with various sensors implemented in the surrounding infrastructure. The infrastructure can share real-time data, such as traffic conditions, road hazards, and weather updates, facilitating safer and more efficient navigation. Within this framework, cooperative sensing is a crucial component, augmenting the onboard sensing capabilities of autonomous vehicles. Cooperative sensing surpasses traditional onboard sensors by leveraging a shared sensor network among vehicles and infrastructure. This approach mitigates challenges posed by occlusion, where objects are obscured from a vehicle's direct view. By pooling information from multiple sources, autonomous vehicles can gain a more comprehensive understanding of their surroundings, leading to enhanced safety and performance on the road. This study addresses a literature gap regarding information flow from real-world scenes to environmental models for cooperative V2I systems. It explores three core concepts essential for understanding the environment: sensing, perception, and mapping. This paper identifies the specific information required from infrastructure nodes, proposes an optimized sensor suite, discusses data processing algorithms, and investigates effective spatial model representations for cooperative sensing. This research informs the reader about the different challenges and opportunities associated with a V2I cooperative sensing system.


  • Open Access

    HDLGen-ChatGPT Case Study: RISC-V Processor VHDL and Verilog Model - Testbench and EDA Project Generation

    by Fearghal Morgan, John Patrick Byrne, Abishek Bupathi, Roshan George, Adnan Elahi, Frank Callaly, Seán Kelly, Declan O'LOughlin

    This paper presents the open source HDLGen-ChatGPT application, working in tandem with ChatGPT-3.5, the free online large language model (LLM) chat interface. The tools enable fast digital systems design and test specification capture, and automatic generation of both VHDL and Verilog models, and testbenches, and AMD Vivado and Intel Quartus Electronic Design Automation (EDA) projects. EDA tools check the generated HDL syntax, simulate and synthesise HDL models, and follow the steps to FPGA hardware prototyping. The tools exploit a formal, top-down design and test specification documentation process, domain knowledge, and the flexibility of LLMs, for HDL code generation. Results are included for a hierarchical RV32I RISC-V processor design. Process steps are illustrated for the RISC-V 32 × 32-bit register bank component. The process typically requires only minimal manual HDL capture or editing, and often none at all. URLs to tutorial videos for the complete RISC-V design are provided on the GitHub project repository. The paper evaluates the results and provides HDLGen-ChatGPT and ChatGPT usage recommendations. The tools can be applied in digital systems training programmes, with reduced emphasis on the assessment of HDL model and testbench capture and generation, while maintaining strong emphasis on the assessment of system design, test planning and documentation, HDL simulation verification/debug, and analysis of synthesised netlists.