1 vin.types.model_inputs

vin.types.model_inputs

Model-input containers shared by VIN scorer implementations.

This module owns dataclasses for intermediate tensors assembled during VIN forward passes. They are separate from helper functions so models, diagnostics, and tests can depend on explicit contracts without importing utility buckets.

1.1 Classes

Name Description
FieldBundle Voxel-aligned scene-field tensors built from actor-visible EVL outputs.
GlobalContext Pose-conditioned global scene context tensors.
PoseFeatures Reference-rig-frame features for VIN candidate views.
PreparedInputs Prepared inputs for VIN v3-style forward passes.

1.1.1 FieldBundle

vin.types.model_inputs.FieldBundle(field_in, field, aux)

Voxel-aligned scene-field tensors built from actor-visible EVL outputs.

1.1.1.1 Attributes

Name Description
field_in Tensor["B C_in D H W", float32] raw scene field.
field Tensor["B C_out D H W", float32] projected scene field.
aux Auxiliary channels, such as normalized counts and occupancy probabilities.

1.1.2 GlobalContext

vin.types.model_inputs.GlobalContext(pos_grid, global_feat)

Pose-conditioned global scene context tensors.

1.1.2.1 Attributes

Name Description
pos_grid Tensor["B 3 D H W", float32] normalized voxel position grid.
global_feat Tensor["B N_q C", float32] pose-conditioned global features.

1.1.3 PoseFeatures

vin.types.model_inputs.PoseFeatures(pose_enc, pose_vec, candidate_center_rig_m)

Reference-rig-frame features for VIN candidate views.

1.1.3.1 Attributes

Name Description
pose_enc Tensor["B N_q E", float32] pose encoder output.
pose_vec Tensor["B N_q D", float32] pose vector fed into the encoder.
candidate_center_rig_m Tensor["B N_q 3", float32] candidate centers in reference-rig metres.

1.1.4 PreparedInputs

vin.types.model_inputs.PreparedInputs(
    pose_world_cam,
    pose_world_rig_ref,
    t_world_voxel,
    batch_size,
    num_candidates,
    device,
    snippet,
)

Prepared inputs for VIN v3-style forward passes.

1.1.4.1 Attributes

Name Type Description
pose_world_cam PoseTW PoseTW["B N_q 12"] world-from-camera candidate poses.
pose_world_rig_ref PoseTW PoseTW["B 12"] reference rig pose in world frame.
t_world_voxel PoseTW PoseTW["B 12"] world←voxel pose for the EVL voxel grid.
batch_size int Batch size inferred from candidate poses.
num_candidates int Number of candidate views per batch item.
device torch.device Device used for tensors in the forward pass.
snippet VinSnippetView VIN snippet view with padded semidense evidence.
Back to top