Data-driven crop growth simulation on time-varying generated images using multi-conditional generative adversarial networks

Table 6 Plant-specific evaluation of mixture biomasses (SW/FB) assessed by MAE and ME in the unit t ha\(^{-1}\) given for all (OA) and mixture (Mix) fields. Each row represents a distinct image generation model trained on a varying combination of conditions time (t), treatment (trt), and simulated biomass (bm); for testing, only the input image and t are varied. Overall fields, MAE is reported for generations with different \(|\Delta t|\) filters: \(\text {T}_0\): identity \(|\Delta t|=0\); ST: short-term \(1\le |\Delta t|\le 10\); LT: long-term \(|\Delta t|\ge 11\)

	Train		OA				OA	Mix	Mix
	conds.		MAE				ME	MAE	ME
	t trt bm		\(\text {T}_0\)	ST	LT	ø	ø	ø	ø
Mixed-CKA	\(\checkmark\) \(\times\) \(\times\)	SW	0.22	0.42	0.39	0.38	0.12	0.31	0.20
Mixed-CKA	\(\checkmark\) \(\times\) \(\times\)	FB	0.16	0.34	0.30	0.28	− 0.12	0.25	− 0.17
Mixed-CKA	\(\checkmark\) \(\checkmark\) \(\times\)	SW	0.30	0.22	0.25	0.24	0.09	0.25	0.15
Mixed-CKA	\(\checkmark\) \(\checkmark\) \(\times\)	FB	0.24	0.16	0.19	0.19	− 0.13	0.24	− 0.15
Mixed-CKA	\(\checkmark\) \(\checkmark\) \(\checkmark\)	SW	0.17	0.21	0.18	0.18	− 0.02	0.18	0.05
Mixed-CKA	\(\checkmark\) \(\checkmark\) \(\checkmark\)	FB	0.11	0.16	0.14	0.13	− 0.01	0.15	− 0.04
Mixed-WG\(^{1}\)	\(\checkmark\) \(\times\) \(\times\)	SW	0.45	1.25	1.14	1.07	0.18	1.06	0.24
Mixed-WG\(^{1}\)	\(\checkmark\) \(\times\) \(\times\)	FB	0.41	0.48	0.67	0.64	− 0.04	0.62	− 0.11

\(^{1}\) Transferability check: Model trained on Mixed-CKA and applied to Mixed-WG

ISSN: 1746-4811