Frontoparallel Motion Tuned:

These are the canonical FP models shown in Figure 5 of the manuscript.

• V5_ME_Binoc.C, the FP component model.
• V5_ME_Binoc.P, the FP pattern model.

These models are the binocular versions of the example component and pattern cell fits from Rust et al., that we showed in Figures 3 and 4 of the manuscript. These use the specfic MT weight distributions from the fits Rust et al. obtained for their examples of a strongly component and pattern-type cell. Strictly monocular versions of these models are also available (see the V5_ME model page).

• V5_ME_Binoc.a, the example cell in Figure 4A and 8A of Rust et al. (2006).
• V5_ME.Binoc.e, the example cell in Figure 4E and 8E of Rust et al. (2006).

3D Motion Tuned:

These are the 3D motion tuned models that produced the results shown in Figures 7 and 8.

• V5_ME_Binoc.3DT.C, the 3DT component model.
• V5_ME_Binoc.3DT.P, the 3DT pattern model.

3D Motion Biased Cell:

The 3D motion biased model we present in Figure 10 is a variation of our canonical FP component model with an added MT imbalance. Here is a link to an edited version of the model file with the additional MT imbalance.

• V5_ME_Binoc.3DB.C, the FP component model with right_eye_w set to 0.5.

Binocular V1 Models:

There are two parameters that adjust binocular integration at the V1 stages of our models:

1. V1 ocular dominance is determined by the parameter binoc_balance. This parameter takes values between 0.5 and 1.0, with a value of 0.5 setting equal mixing (50/50) of left and right eye signals for all V1 channels, and a value of 1.0 setting the V1 channels to be purely monocular (this is the default value).
2. The order of operations for motion opponency and binocular integration in V1 is set by the parameter early_mix_stage. This can take values of 'linear', 'normalized' or 'opponent', with binocular integration occurring after the indicated stage, or 'none' indicating no V1 binocular mixing (default is 'none').

Here we provide FP and 3DT versions of component and pattern models with binoc_balance set to 0.5 and early_mix_stage set to 'opponent'. Results from only the pattern models are shown in Figure 11 of the manuscript.

• V5_ME_Binoc.V1OB.C, FP component model.
• V5_ME_Binoc.V1OB.P, FP pattern model.
• V5_ME_Binoc.3DT.V1OB.C, 3DT component model.
• V5_ME_Binoc.3DT.V1OB.P, 3DT pattern model.

Binocular Disparity Models:

There are two parameters controlling computation of binocular disparity at the V1 binocular integration stage of our models:

1. The parameter disparity_flag turns disparity computation on (1) or off (0, default).
2. The parameter disparity_norm sets the signals used for normalization of the monocular V1 signals prior to disparity computation (default is 1 - motion energy)

Here we provide FP and 3DT versions of component and pattern models with binocular disparity shown in Figure 12 of the manuscript.

• V5_ME_BDE.C, FP component model.
• V5_ME_BDE.P, FP pattern model.
• V5_ME_BDE.3DT.C, 3DT component model.
• V5_ME_BDE.3DT.P, 3DT pattern model.