Source: Ghana Web
In a remarkable display of innovation and determination, Ghana has witnessed groundbreaking efforts in the fight against desertification. Armed with cutting-edge AI technology, the transformative project was spearheaded by the visionary Damoah brothers, Dominic and Michael, alongside their esteemed professor, Mr. JK Arthur, all experts in the field of AI and Machine Learning. In this exclusive feature, we delve into their pivotal roles in the project and the transformative potential of this AI-based platform.
Dominic and Michael Damoah, in collaboration with their professor Mr. JK Arthur, embarked on a visionary research project. Their mission was to leverage the power of artificial intelligence to address one of Ghana’s most pressing environmental challenges – desertification. This phenomenon, exacerbated by illicit mining activities and other detrimental factors driven by the relentless pursuit of financial gain, posed a dire threat to the nation’s ecosystems.
“The reason why we did this was to help the struggle and fight to preserve natural resources in Ghana. There are already a lot of things that are happening to our natural resources due to Galamsey and so many others that I would call them crimes. So many other negative things that are happening to our environment now in the pursuit of financial gain,” Dominic Damoah, the eldest of the two brothers said.
At the heart of their groundbreaking initiative was their AI-based platform, utilizing image classification to identify desertification-affected regions with exceptional precision. By ingeniously deploying satellite imagery and drone-captured aerial photos, the platform generated invaluable insights.
Explaining how this works, Dominic said, “The plan for deployment was that we would send up a drone every now and then into areas that we know have that kind of activity. And that drone would basically take aerial photos. And on a periodic basis, maybe on a three-week basis, those photos would come into, let’s say, a command center we envisage.”
To fuel this ambitious endeavor, the trio meticulously optimized the Aerial Images dataset, a public repository tailored for training AI models. They adopted the acclaimed VGG 19 pre-trained model for aerial image classification as the foundation of their methodology. Under their collective guidance, they took it a step further, honing and optimizing the model to achieve unparalleled accuracy and precision.
Yet, the relentless pursuit of innovation did not stop there. Determined to push the boundaries, the Damoah brothers and Mr. Arthur, working collaboratively, devised an ingenious transfer learning technique. This involved integrating an additional classification network that drew from the VGG 19 model’s weights, elevating their platform to an unprecedented level of refinement.
In a step-by-step account of how the duo created the project, Micahel Damoah explained;
“So basically, it’s an AI project. And with every AI project, you need a data set to build your model around.
“So what we did was, we went online and found a data set that is; the Aerial Images data set (AID). It’s a public data set which contains a bunch of aerial images that you can use to train a model for whichever purpose.
“In our case here, it’s detecting desertification. So then what we further went on to do, is to adopt the VGG 19 pre-trained model for aerial image classification. This is a very popular artificial intelligence model in the image classification field.”
He continued, “Then we optimized this VGG 19 model for better accuracy and efficiency with its predictions. Then, on top of the VGG 19 model, we wrote an extra classification network that takes its weights from the VGG 19 model to better refine the output.
The model that we created was built with a TensorFlow Keras API. And what it basically does is, the top model uses the VGG pretrained network for feature extraction. So it’s essentially a classic case of transfer learning.”
“What’s noteworthy is that Mr. JK Arthur played a pivotal role in drawing out the architectural design for this groundbreaking scientific experiment, providing invaluable insights into structuring the project. With a wealth of experience in environmental science and Artificial Intelligence, Mr. Arthur’s contributions were instrumental in shaping the foundations of our desertification detection AI platform.
His expertise ensured that our model not only achieved high accuracy but also considered critical environmental factors and sustainability aspects, making it a powerful tool for monitoring and combating desertification on a broader scale.
Mr. Arthur’s guidance in designing a robust and adaptable architecture laid the groundwork for a system capable of addressing the complex challenges posed by desertification, marking a significant leap forward in environmental AI research.
The true stroke of genius came when the team transformed their AI masterpiece into a web application. This innovation meant that environmental protection agencies, the Ghanaian government, and organizations across West Africa and beyond could now harness the power of their platform with ease. The intuitive web interface empowered users to test and verify the platform’s capabilities, reinforcing its credibility as an indispensable tool in the fight against desertification.
Michael Damoah has showcased the web application’s capabilities, highlighting its potential impact on Ghana’s environment and vulnerable regions worldwide. With a seamless integration of human intellect and artificial intelligence, the platform promised to expedite the identification of regions severely affected by desertification, hastening the implementation of much-needed remedial measures.
This profound commitment to safeguarding Ghana’s natural heritage, coupled with unmatched technical expertise, positioned the Damoah brothers and Mr. JK Arthur as pioneers in AI-driven desertification detection in West Africa. Their platform’s potential to revolutionize environmental conservation captured the attention of global environmentalists and conservationists, setting a precedent for sustainable development worldwide.
As we delve into the story of Dominic and Michael Damoah, it becomes evident that their efforts transcend mere scientific inquiry. Theirs is a tale of relentless passion, visionary thinking, and an unwavering dedication to addressing one of the most pressing challenges of our time.
Their journey represents a beacon of hope for Ghana and the entire world grappling with ecological upheaval.
In conclusion, the collaborative efforts of the Damoah brothers and Mr. JK Arthur, in developing this AI-based platform, exemplify the transformative potential of technology harnessed for the greater good. As we embrace this remarkable advancement, the hope is that governments, environmental organizations, and conservationists worldwide will rally behind this collaborative vision, joining hands in pursuing a sustainable and resilient planet for generations to come.
“Basically it’s just my hope that the environmental protection agencies and probably the government of Ghana and other governments out here adopt this in the fight against desertification out here in Ghana, and across West Africa and probably other parts of the world,” Michael stressed.
Source: Ghana Web