Research map / Image generation / Text-to-image diffusion
Toward Deeper Emotional Reflection: Crafting Affective Image Filters With Generative Priors
Peixuan Zhang, Shuchen Weng, Jiajun Tang, Si Li, Boxin Shi
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 |
|---|---|---|---|
| add content preservation module (ablation) | AIF-D · AIF dataset test set | SSIM (%): 13.48 -> 57.74 | Improved |
| add in-depth emotional reasoning with LLM cues (ablation) | AIF-D · AIF dataset test set | EAcc (%): 35.87 -> 37.48 | Improved |
| add voting ensemble mechanism (ablation) | AIF-D · AIF dataset test set | EAcc (%): 34.18 -> 37.48 | Improved |
| add emotional reflection decoder instead of vanilla decoder (ablation) | AIF-D · AIF dataset test set | SG (per mille): 1.4639 -> 1.3416 | Improved |
| add texture mapping loss instead of standard style loss (ablation) | AIF-D · AIF dataset test set | SSD: 1.2824 -> 1.1123 | Improved |
| add emotional reflection decoder (ablation) | AIF-D · AIF dataset test set | EAcc (%): 35.21 -> 37.48 | Improved |
| switch backbone from transformer AIF-B to diffusion-based AIF-D | AIF-B · AIF dataset test set | SSIM (%): 56.15 -> 57.74 | Improved |
| switch from AIF-B to AIF-D | AIF-B · AIF dataset test set | EAcc (%): 36.02 -> 37.48 | Improved |
| switch from AIF-B to AIF-D | AIF-B · AIF user study (AMT, 200 samples) | EPS user study (%): 18.52 -> 39.51 | Improved |
| set style/content balance gamma to 0.3 instead of 0.1 (ablation) | AIF-D · AIF dataset test set | EAcc (%): 36.64 -> 37.48 | Improved |
| set gamma to 0.3 instead of 3.0 (ablation) | AIF-D · AIF dataset test set | SSIM (%): 55.11 -> 57.74 | Improved |
| set gamma to 0.3 instead of 1.0 (ablation) | AIF-D · AIF dataset test set | SSD: 1.1679 -> 1.1123 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- CLIPstyler — CLIPstyler: Image Style Transfer With a Single Text Condition (CVPR 2022) · better on 5 of 5 shared results
- DiffusionCLIP — DiffusionCLIP: Text-Guided Diffusion Models for Robust Image Manipulation (CVPR 2022) · better on 5 of 5 shared results
- ManiGAN — ManiGAN: Text-Guided Image Manipulation (CVPR 2020) · better on 5 of 5 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- EmoEdit: Evoking Emotions through Image Manipulation (CVPR 2025 · similarity 0.86)
- Customizing Visual Emotion Evaluation for MLLMs: An Open-vocabulary, Multifaceted, and Scalable Approach (ICLR 2026 · similarity 0.85)
- EmoGen: Emotional Image Content Generation with Text-to-Image Diffusion Models (CVPR 2024 · similarity 0.85)
- EEmo-Logic: A Unified Dataset and Multi-Stage Framework for Comprehensive Image-Evoked Emotion Assessment (ICML 2026 · similarity 0.85)
- Multiple Feature Refining Network for Visual Emotion Distribution Learning (AAAI 2025 · similarity 0.84)
- Bridge Then Begin Anew: Generating Target-Relevant Intermediate Model for Source-Free Visual Emotion Adaptation (AAAI 2025 · 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.
