I worked on improving architecture backbones and developing ensembles of vision models, including CoatNet, CBAM, YOLO, Transformers, and Retention Networks. My focus was enhancing object detection rates in underwater imagery and plankton classification using shadowgraph imagery captured with line-scan cameras. Given their critical role in the ecosystem, plankton contribute to over half of the earth’s primary productivity, forming the foundation of aquatic food webs and global carbon cycles. This research holds significance in understanding environmental health and sustaining vital ecosystems.