Table 2.

Optimized parameters of the model

Model	Input size	Batch size	Epoch	Optimizer	Learning rate	Weight decay	Momentum
FISH-Net	320 × 320	64	30	Adam	0.01	0.0001	0.9
YOLO-V7s	320 × 320	32	100	SGD	0.01	0.0001	0.9
CenterNet2	320 × 320	32	50	SGD	0.01	0.0001	0.9
Multi-scale MobileNet-YOLO-V4	416 × 416	24	200	SGD	0.001	0.0001	0.9
CACNET	512 × 512	1	1000	Adam	0.01	0.001	0.8