Tomoya Nishino

The paper standardizes AI-derived fish length-frequency distributions using bin-specific error matrices to improve comparability across surveys.

This paper develops an image-processing method to estimate curved body length of tuna larvae from microscope images without training, achieving substantial agreement with manual measurements for larvae ≤5.0 mm.

This study evaluates how varying input image sizes affects fish species identification accuracy using deep learning, finding that larger images improve accuracy, especially for underperforming species.