The NYBG has presented us with a unique and complex challenge: a significant portion of the over 40 million digitized images in botanical databases are classified as "non-standard," including images of animals or colored plant illustrations rather than the actual pressed plant specimens needed for research. These anomalies pose a substantial barrier to the application of machine learning in botanical studies, such as species recognition and analyzing biodiversity trends. Our team's task is to develop a sophisticated model capable of accurately identifying and classifying these images to facilitate the curation of datasets for environmental research.
This project is more than a technical endeavor; it represents a critical contribution to environmental science. By automating the identification of non-standard images, our team will significantly streamline the process of dataset curation for researchers. This breakthrough has the potential to accelerate environmental studies, offering new insights into biodiversity, conservation, and the effects of climate change. We're not just developing a machine learning model; we're contributing to a global effort to understand and protect our planet's plant diversity.
Our primary objective is to employ advanced image classification techniques to create a model that distinguishes between various classes of images with high accuracy. Success in this venture could lead to the development of a publicly available tool by NYBG, which would enable scientists worldwide to efficiently filter non-standard images from their datasets. This tool would be instrumental in conducting pivotal biodiversity research with the aid of machine learning.
By participating in this project, our team is at the forefront of merging technology with environmental science. We're not just honing our skills in machine learning; we're paving the way for significant discoveries that will help understand and mitigate the impacts of human activity on plant diversity. Together, we're using our knowledge and expertise to contribute to a greener, more sustainable future.