Research map / Image generation / Text-to-image diffusion
The Silent Assistant: NoiseQuery as Implicit Guidance for Goal-Driven Image Generation
Ruoyu Wang, Huayang Huang, Ye Zhu, Olga Russakovsky, Yu Wu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| select aligned noise from library (NoiseQuery) | SD 1.5 · DrawBench | ImageReward: 0.04 -> 0.08 | Improved |
| select aligned noise from library (NoiseQuery) | SD-Turbo · MSCOCO | CLIPScore: 31.51 -> 31.76 | Improved |
| add NoiseQuery on top of Diffusion-DPO | SD 1.5 + Diffusion-DPO · DrawBench | ImageReward: 0.09 -> 0.17 | Improved |
| add NoiseQuery on top of CFG++ | SD 2.1 + CFG++ · DrawBench | ImageReward: 0.12 -> 0.27 | Improved |
| add NoiseQuery on top of ReNO | SD-Turbo + ReNO · DrawBench | ImageReward: 1.67 -> 1.71 | Improved |
| add NoiseQuery on top of LaVi-Bridge | PixArt-alpha + LaVi-Bridge · DrawBench | ImageReward: 0.63 -> 0.72 | Improved |
| increase noise library size from 0.5k to 100k (ablation) | NoiseQuery · MSCOCO | CLIPScore: 31.51 -> 31.74 | Improved |
| transfer library built on SD v2.1 zero-shot to PixArt-alpha (ablation) | NoiseQuery · MSCOCO | CLIPScore: 31.48 -> 31.67 | Improved |
| transfer library built on SD v2.1 zero-shot to SD v1.4 (ablation) | NoiseQuery · MSCOCO | PickScore: 21.40 -> 21.47 | Improved |
| use library built with DDIM under PDNM scheduler (ablation) | SD 2.1 · DrawBench | CLIPScore: 30.91 -> 31.67 | Improved |
| use DDIM-built library with DPM++ 2M SDE Karras scheduler (ablation) | SD 2.1 · DrawBench | HPS v2: 23.32 -> 25.68 | Improved |
| use DDIM-built library with Euler scheduler (ablation) | SD 2.1 · DrawBench | PickScore: 21.35 -> 21.38 | Improved |
Closest papers by meaning
- InitNO: Boosting Text-to-Image Diffusion Models via Initial Noise Optimization (CVPR 2024 · similarity 0.89)
- Noise Diffusion for Enhancing Semantic Faithfulness in Text-to-Image Synthesis (CVPR 2025 · similarity 0.88)
- Ranni: Taming Text-to-Image Diffusion for Accurate Instruction Following (CVPR 2024 · similarity 0.86)
- Scaling Rectified Flow Transformers for High-Resolution Image Synthesis (ICML 2024 · similarity 0.86)
- Rethinking Training for De-biasing Text-to-Image Generation: Unlocking the Potential of Stable Diffusion (CVPR 2025 · similarity 0.86)
- Adjusting Initial Noise to Mitigate Memorization in Text-to-Image Diffusion Models (NeurIPS 2025 · similarity 0.86)
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.
