Skip to content

T3B — Depth estimation — photometric shift

Paper evidence: main.pdf · Block findings

Lemma: D3 · Stack: pytorch Nuisance key: augmentation

Production change: Lighting/texture shift; depth target meaning unchanged.

Notebook (Run All, built-in demo): t03b-depth-augmentation.ipynb

pip install matching-pmh torch
# Open the notebook and Run All

What this task achieved (headline)

E1_aniso beats E1 on hard photometric depth stress (combined_hard AbsRel 0.2152 vs 0.2191).

baseline E1 E1_aniso
0.2033 AbsRel 0.1951 0.2152 (photo hard)

Paper preset: t3b_depth_d3 · from pmh.benchmark.presets import get_preset

Subtasks (paper)

NYU depth pipeline (train + calibrate)

E1_aniso wins photometric hard stress.

Preset: t3b_depth_d3

Photometric subspace calibration

Aug-delta Gram rank 32.

Preset: t3b_depth_d3

E1_wrong control (random W)

Strengthening / falsification arm.

Run with matching-pmh

from pmh import PMHTrainer, evaluate_robust_fit
# nuisance="augmentation"

Do not use PMH when

Different depth semantics or scale definition at deploy.

Replace demo data with yours

Swap demo loaders for your train_loader, source_batches, target_batches, and deploy holdout. Hook the backbone before your task head.

← All 13 tasks · Quickstart