Emmanuel Chibuikem Nnadozie

Mr. Emmanuel Chibuikem Nnadozie

Mechatronic Engineering, University of Nigeria, Nsukka, Enugu, Nigeria
orcid

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
28 April 2023 in Remote Sensing
Verified
Journal Article
Open access
PDF file
ENEmmanuel Chibuikem Nnadozie
Emmanuel Chibuikem Nnadozie
OIOgechukwu N. Iloanusi
Ogechukwu N. Iloanusi
OAOzoemena Ani
Ozoemena Ani
Kang YuKang Yu
Kang Yu
A significant number of object detection models have been researched for use in plant detection. However, deployment and evaluation of the models for real-time detection as well as for crop counting under varying real field conditions is lacking. In this work, two versions of a state-of-the-art object detection model—YOLOv5n and YOLOv5s—were deployed and evaluated for cassava detection. We compared the performance of the models when trained with different input image resolutions, images of different growth stages, weed interference, and illumination conditions. The models were deployed on an NVIDIA Jetson AGX Orin embedded GPU in order to observe the real-time performance of the models. Results of a use case in a farm field showed that YOLOv5s yielded the best accuracy whereas YOLOv5n had the best inference speed in detecting cassava plants. YOLOv5s allowed for more precise crop counting, compared to the YOLOv5n which mis-detected cassava plants. YOLOv5s performed better under weed interference at the cost of a low speed. The findings of this work may serve to as a reference for making a choice of which model fits an intended real-life plant detection application, taking into consideration the need for a trade-off between of detection speed, detection accuracy, and memory usage.

4 November 2023 in Multimedia Tools and Applications
Verified
Journal Article
Open access
PDF file
ENEmmanuel Chibuikem Nnadozie
Emmanuel Chibuikem Nnadozie
PHPablo Casaseca de la Higuera
Pablo Casaseca de la Higuera
OIOgechukwu Iloanusi
Ogechukwu Iloanusi
OAOzoemena Ani
Ozoemena Ani
CACarlos Alberola-López
Carlos Alberola-López
Deep learning-based object detection models have become a preferred choice for crop detection tasks in crop monitoring activities due to their high accuracy and generalization capabilities. However, their high computational demand and large memory footprint pose a challenge for use on mobile embedded devices deployed in crop monitoring settings. Various approaches have been taken to minimize the computational cost and reduce the size of object detection models such as channel and layer pruning, detection head searching, backbone optimization, etc. In this work, we approached computational lightening, model compression, and speed improvement by discarding one or more of the three detection scales of the YOLOv5 object detection model. Thus, we derived up to five separate fast and light models, each with only one or two detection scales. To evaluate the new models for a real crop monitoring use case, the models were deployed on NVIDIA Jetson nano and NVIDIA Jetson Orin devices. The new models achieved up to 21.4% reduction in giga floating-point operations per second (GFLOPS), 31.9% reduction in number of parameters, 30.8% reduction in model size, 28.1% increase in inference speed, with only a small average accuracy drop of 3.6%. These new models are suitable for crop detection tasks since the crops are usually of similar sizes due to the high likelihood of being in the same growth stage, thus, making it sufficient to detect the crops with just one or two detection scales.

17 November 2022 in bioRxiv
Verified
Preprint
Open access
PDF file
ENEmmanuel Chibuikem Nnadozie
Emmanuel Chibuikem Nnadozie
OIOgechukwu Iloanusi
Ogechukwu Iloanusi
OAOzoemena Ani
Ozoemena Ani
Kang YuKang Yu
Kang Yu
Most deep learning-based weed detection methods either yield high accuracy, but are slow for real-time applications or too computationally intensive for implementation on smaller devices usable on resource-constrained platforms like UAVs; on the other hand, most of the faster methods lack good accuracy. In this work, two versions of the deep learning-based YOLOv5 object detection model – YOLOv5n and YOLOv5s - were evaluated for cassava detection as a step towards real-time weed detection. The performance of the models were compared when trained with different image resolutions. The robustness of the models were also evaluated under varying field conditions like illumination, weed density, and crop growth stages. YOLOv5s showed the best accuracy whereas YOLOv5n had the best inference speed. For similar image resolutions, YOLOv5s performed better, however, training YOLOv5n with higher image resolutions could yield better performance than training YOLOv5s with lower image resolutions. Both models were robust to variations in field conditions. The speed vs accuracy plot highlighted a range of possible speed/accuracy trade-offs to guide real-time deployment of the object detection models for cassava detection.


27 July 2023
Verified
Book Chapter
Closed access
ENEmmanuel Chibuikem Nnadozie
Emmanuel Chibuikem Nnadozie
OIO.N. Iloanusi
O.N. Iloanusi
OAOzoemena Ani
Ozoemena Ani
Kang YuKang Yu
Kang Yu
Precision agriculture is a reality in agriculture and is playing a key role as the industry comes to terms with the environment, market forces, quality requirements, traceability, vehicle guidance ...

Verified
Journal Article
Open access
PDF file
ENEmmanuel Chibuikem Nnadozie
Emmanuel Chibuikem Nnadozie
SMSusana Merino-Caviedes
Susana Merino-Caviedes
Daniel-Antonio de Luis RomanDaniel-Antonio de Luis Roman
Daniel-Antonio de Luis Roman
MMMarcos Martin-Fernandez
Marcos Martin-Fernandez
CACarlos Alberola-López
Carlos Alberola-López
High-performance medical segmentation models are often benchmarked on high-end GPUs. Such benchmarks do not provide useful performance insights for point-of-care low-end devices. This work, firstly, posits that to achieve improved clinical adoption of AI-powered segmentation models, especially in reduced manpower settings like rural hospitals, we need benchmarks that provide actionable insights on the degree to which high-performance models address five deployment constraints viz: resource-effectiveness for low-end computing devices, clinically acceptable accuracy, clinically compatible execution times, localization of user data, and user-based finetuning. In this work, five state-of-the-art foundation segmentation models and one target-specific model were systematically evaluated on three multi-organ medical datasets. Furthermore, the best-ranking foundation model and target-specific model were benchmarked on three low-end devices. Our findings show that lightweight foundation models provided the best performance trade-off and are easily user-fine-tuned on custom datasets. Target-specific models provide high accuracy out-of-the-box, but may require significant optimisation to deliver comparably fast execution times and user-based finetuning on low-end devices. The methods and results from this research provide actionable insights on high-performance medical segmentation models for low-end computing devices, as a necessary step towards improved adoption in resource-limited clinical settings.