Research map / Image generation / Autoregressive and latent generation
Generalized rectifier wavelet covariance models for texture synthesis
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Replace classical PS wavelet statistics with ALPHA_I rectified-wavelet covariance statistics | Texture synthesis · Gray-scale and color textures | visual quality | Helped |
| Use ALPHA_I rectified-wavelet covariance statistics instead of random-filter statistics | Texture synthesis · Gray-scale and color textures | visual quality | No clear effect |
| Add correlations between spatially shifted rectified wavelet coefficients across all coefficient pairs (ablation) | ALPHA_S · Gray-scale textures and non-stationary images | reproduction of salient shapes and contours | Helped |
| Add correlations across a wider range of scales (ablation) | ALPHA_S · Gray-scale textures | visual quality | Helped |
| Add a second phase-shift index to the ALPHA_I correlations (ablation) | ALPHA_I, producing ALPHA_L · Gray-scale textures | diversity | Hurt |
| Add a second phase-shift index to the ALPHA_I correlations (ablation) | ALPHA_I, producing ALPHA_L · Gray-scale textures | geometric information captured | Helped |
| Restrict non-zero spatial shifts to correlations within the same color channel (ablation) | ALPHA_C · Color textures | visual quality | No clear effect |
| Use Morlet wavelets instead of Simoncelli steerable wavelets (ablation) | ALPHA_I · Texture synthesis | contour reproduction | Helped |
| Remove the non-zero spatial-shift correlation constraint (ablation) | PS · Binary non-stationary images | reproduction of the observation | Hurt |
Closest papers by meaning
- Texture Interpolation for Probing Visual Perception (NeurIPS 2020 · similarity 0.88)
- Visualizing and Understanding Deep Texture Representations (CVPR 2016 · similarity 0.88)
- Diversified Texture Synthesis With Feed-Forward Networks (CVPR 2017 · similarity 0.87)
- Wavelet-Driven Masked Image Modeling: A Path to Efficient Visual Representation (AAAI 2025 · similarity 0.85)
- Diversify, Contextualize, and Adapt: Efficient Entropy Modeling for Neural Image Codec (NeurIPS 2024 · similarity 0.85)
- Generating Non-Stationary Textures using Self-Rectification (CVPR 2024 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.