Working with SoftBank Corp. and Nosan Corporation, NeuralX, Inc. published a scientific paper at SIGGRAPH Asia 2021, a top tier international conference in the field of computer graphics and interactive technologies. NeuralX presented the paper at the Tokyo International Forum in Tokyo on December 14th, 2021.
The paper details a new approach to fish counting (a), leveraging a synthetic dataset to train a sophisticated computer vision algorithm. The proposed system is expected to make contributions in smart fish farming, automating fish counting and sizing. This accuracy enables optimizations for the feeding and operations of the fish farm, as well as strategic insights for long-term planning.
“The paper is the culmination of efforts that combine advances in computer vision, bio-inspired simulations, and synthetic data. Our computer vision algorithm counts with 97% accuracy, providing insights for the efficient growth and feeding of farm-raised fish.”
– Dr. Masaki Nakada of NeuralX
An estimated 3 billion rely on seafood as their primary source of protein. As the world’s population grows to 10 billion people, the need to support this demand becomes a necessity. The current process (b) for counting fish is quite manual in nature. This time-consuming task requires lifting the fish out of the water and weighing them on scales. In addition, it can be quite traumatic for the fish and unnecessarily stress or kill them.
SIGGRAPH (Special Interest Group on Computer Graphics and Interactive Techniques) is an annual conference on computer graphics (CG) organized by the ACM SIGGRAPH, starting in 1974. The main conference is held in North America; SIGGRAPH Asia, a second conference held annually, has been held since 2008 in countries throughout Asia.