Research map / Image generation / Editing and style transfer
CompBench: Benchmarking Complex Instruction-guided Image Editing
Bohan Jia, Wenxuan Huang, Yuntian Tang, Junbo Qiao, Jincheng Liao, Shaosheng Cao, Fei Zhao, Zhaopeng Feng, Zhouhong Gu, Zhenfei Yin, Lei Bai, Wanli Ouyang and 5 more
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 |
|---|---|---|---|
| integrate an MLLM for instruction understanding | image editing models · CompBench overall | overall score: top models (Bagel 4.76, Qwen-Image-Edit 4.32) mostly use MLLMs | Improved |
| make the edit a second turn instead of first | all editing models · CompBench multi-turn | background SSIM: Bagel 0.949 -> 0.897; InstructPix2pix 0.682 -> 0.573 | Hurt |
| use complex scenes instead of simple benchmarks | InstructPix2pix · reasoning edits vs ReasonEdit | PSNR / SSIM / CLIP: about -2.5 PSNR, -0.02 SSIM, -0.4 CLIP-Score | Hurt |
| compare unified generation-understanding model against other models | Bagel · CompBench | top metrics: top results in 18 out of 37 metrics | Improved |
| use viewpoint editing instead of local editing | editing models · CompBench viewpoint editing | score (0-10): best avg 5.615 (Qwen-Image-Edit); most others below 3 | Hurt |
| add reasoning data or chain-of-thought | SmartEdit / GoT · CompBench | qualitative | Improved |
Closest papers by meaning
- Envisioning Beyond the Pixels: Benchmarking Reasoning-Informed Visual Editing (NeurIPS 2025 · similarity 0.89)
- I2EBench: A Comprehensive Benchmark for Instruction-based Image Editing (NeurIPS 2024 · similarity 0.89)
- InsightEdit: Towards Better Instruction Following for Image Editing (CVPR 2025 · similarity 0.88)
- SmartEdit: Exploring Complex Instruction-based Image Editing with Multimodal Large Language Models (CVPR 2024 · similarity 0.88)
- Balancing Preservation and Modification: A Region and Semantic Aware Metric for Instruction-Based Image Editing (ICML 2025 · similarity 0.88)
- KRIS-Bench: Benchmarking Next-Level Intelligent Image Editing Models (NeurIPS 2025 · similarity 0.88)
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.
