Table 1.

Quantitative evaluation of the reconstructions using Yanagi’s and our MSFA (see Figure 6) on synthetic datasets. Note that the bold font indicates the minimal error.

	MSFA	Dataset
		CAVE		Harvard		ICVL
		RMSE	PSNR	RMSE	PSNR	RMSE	PSNR
Noisy (SNR = ∞)	Ours	0.172	21.38	0.165	22.97	0.215	19.80
	Yanagi’s	0.203	20.67	0.190	21.18	0.268	18.42
Noisy (SNR ≈ 30 db)	Ours	0.198	20.19	0.184	21.26	0.244	19.15
	Yanagi’s	0.231	19.39	0.217	19.90	0.293	17.43