ZeroHour
Hugging Face daily paperspublished ()ingested Jiayin Chen, Yicheng Xu, Muting Wang

Mi-Ripple: Restoring Images Degraded by Iterative AI Editing

infoAI researchimportance 25
AI summary · glm-5.3-flash

Mi-Ripple is a diagnosis-guided restoration workflow that removes digital ripple artifacts introduced by iterative AI image editing while preserving structure.

Iterative reference-conditioned image editing can introduce grid-like and granular textures known as digital ripple. Mi-Ripple separates periodic lattice artifacts from content-entangled granular texture, then applies selective spectral notching, structure-aware smoothing, and cleaned-reference regeneration. In fourteen notch-only executions, whole-image residual standard deviation was 0.08-0.44 in CIELAB lightness units, and reference cleaning reduced output debris density by 45% in a paired example.

  • Separates periodic lattice artifacts from granular texture for targeted filtering
  • Combines spectral notching with cleaned-reference regeneration to avoid erasing detail
  • Reference cleaning cut debris density by 45% in a paired test
  • Residual artifact deviation of 0.08-0.44 CIELAB units across trials
ProductsMi-Ripple
AI modelsMi-Ripple
Full article113 words · extracted from huggingface.co · click to collapse

Iterative reference-conditioned image editing can introduce grid-like and granular textures, commonly described as digital ripple. We present Mi-Ripple, a diagnosis-guided restoration workflow that suppresses this digital ripple while protecting image structure. Mi-Ripple separates periodic lattice artifacts from content-entangled granular texture, then combines selective spectral notching, structure-aware smoothing, and cleaned-reference regeneration. This separation enables low-distortion filtering when artifacts are spectrally isolated and visual reconstruction when filtering would erase legitimate detail. Across fourteen notch-only executions, whole-image residual standard deviation is 0.08--0.44 in CIELAB lightness units. In a paired regeneration example, reference cleaning reduces output debris density by 45\%. Mi-Ripple links measurable artifact reduction to visibly cleaner generated images, rather than optimizing a spectral score alone.

Text extracted automatically; images, tables and formatting may be missing. Original: https://huggingface.co/papers/2609.11317