Image Inpainting (masked region recovery) L1-384
Unclaimed Principle — open for contribution
This Principle is declared in the catalog but has no reference solver, no pinned dataset, and is not registered on-chain. There is no reward pool. Submitting a cert against this Principle today will record the cert for reproducibility but pay zero PWM.
To claim it as a Bounty #7 contribution: open a PR adding (1) a reference solver, (2) ≥1 dataset pinned to IPFS, (3) updates to the L3 manifest with dataset CIDs. After verifier-agent triple-review, the founders' 3-of-5 multisig signs PWMRegistry.register() and the Principle becomes mineable.
Forward model E
A binary mask M removes pixels from the image x (unknown pixels replaced by a sentinel); the task recovers x in the masked region using surrounding context and a natural-image prior.
L-DAG
Well-posedness W
- Existence:
- true
- Uniqueness:
- false
- Stability:
- conditional
- κ:
- 10000
Strictly underdetermined inside mask; uniqueness is prior-induced (TV, non-local similarity, learned). Stability decreases super-linearly with mask_ratio and hole diameter.
Solvability C
- Solver class:
- PDE diffusion (Bertalmio), exemplar-based (Criminisi), total-variation (TV), patch-match, learned (DeepFill, partial-conv, LaMa, MAT)
- Convergence rate q:
- 2
- Complexity:
- PDE O(H*W*iterations); patch-match O(H*W*P); learned single forward O(H*W*C*F)