Fig. 8

Comparison of qualitative results between VMUnet-MSADI and the existing models on the corn kernel segmentation task. To better visualize the differences between segmentation predictions and ground truths, we highlight the key region with appropriate boxes.