An Improved EfficientNetV2 Model Based on Visual Attention Mechanism: Application to Identification of Cassava Disease

Algorithm 1

Dynamic attenuation of learning rate.

	Require: epoch, lrf ,
	Require.
	Output: Communities.
(1)	while not converged do
(2)	if t 20 then
(3)	p
(4)	break
(5)	end if
(6)	if 20 < t 60 then
(7)	p
(8)	break
(9)	end if
(10)	if 60 < t 100 then
(11)	p
(12)	break
(13)	end if
(14)	if 100 < t 120 then
(15)	p
(16)	break
(17)	end if
(18)	t + 1
(19)	end while
(20)	return P