Creating Domain-Specific Datasets for Intelligent Environmental Feature Comparison
Coastal environments are dynamic and ecologically significant, yet monitoring across multiple sites and analysis remain challenging due to the lack of domain-specific datasets tailored to their unique features. General-purpose models, including those used for scene graph generation, often fail to c...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
LibraryPress@UF
2025-05-01
|
| Series: | Proceedings of the International Florida Artificial Intelligence Research Society Conference |
| Online Access: | https://journals.flvc.org/FLAIRS/article/view/138913 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Coastal environments are dynamic and ecologically significant, yet monitoring across multiple sites and analysis remain challenging due to the lack of domain-specific datasets tailored to their unique features. General-purpose models, including those used for scene graph generation, often fail to capture the semantic details necessary for meaningful comparisons in this context. This paper outlines the process of creating a domain-specific dataset for coastal environments, focusing on the challenges posed by crowdsourced imagery, such as variability in image sizes, lighting conditions, and camera quality. By leveraging scene graph generation to capture semantic meaning, this research seeks to create a domain-specific dataset suitable for the comparison of coastal environments. This work demonstrates how domain-specific datasets can drive innovation in computer vision and semantic understanding, contributing to the broader field of artificial intelligence by bridging the gap between generalized tools and specialized applications. Ultimately, this effort lays the groundwork for future planned research to develop a pipeline capable of generating comparison metrics based on the semantic content of scenes. Using raw standardized images of coastal environments from the Coastie Initiative, this pipeline aims to go beyond superficial appearance comparisons, offering more meaningful analyses that could enhance our understanding and support conservation efforts.
|
|---|---|
| ISSN: | 2334-0754 2334-0762 |