cutcutcodec.core.analysis.video.quality.vmaf_torch.adm.ADM

class cutcutcodec.core.analysis.video.quality.vmaf_torch.adm.ADM(NEG=False)

Initialize internal Module state, shared by both nn.Module and ScriptModule.

adm_cm(R_h, R_v, R_d, csf_A_h, csf_A_v, csf_A_d, width, height, scale)

Compute numerator for eq. (19) Perform Contrast Masking on restored image R, and spatial and subband Minkowski pooling

adm_csf(A_h, A_v, A_d, width, height, scale)

Apply Contrast Sensitivty Function (CSF) on additive impairments image A

adm_csf_den(O_h, O_v, O_d, width, height, scale)

Compute denominator for eq. (19). Perform CSF weighting (without CM) on original image, and spatial and subband Minkowski pooling

adm_decouple(O_h, O_v, O_d, T_h, T_v, T_d, width, height, scale)

Compute Restored image and Additive impairment image, eq. (12), (15)

adm_dwt2(X)

Compute discrete wavelet transform

adm_num_den(X_ref, X_dst)

Compute ADM numerator and denominator for several scales

forward(X_ref, X_dst)

Define the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.