Research map / Image generation / Editing and style transfer
MCIE: Multimodal LLM-Driven Complex Instruction Image Editing with Spatial Guidance
Xuehai Bai, Xiaoling Gu, Akide Liu, Hangjie Yuan, YiFan Zhang, Jack Ma
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 spatial-aware cross-attention (SACA) (ablation) | MCIE-E1 · CIE-Bench | IC: 0.4922 -> 0.7932 | Improved |
| add background-consistent cross-attention (BCCA) to SACA (ablation) | MCIE-E1 · CIE-Bench | BC: 0.6851 -> 0.7831 | Improved |
| add BCCA to SACA (ablation) | MCIE-E1 · CIE-Bench | CLIP-I: 0.8566 -> 0.8899 | Improved |
| add BCCA to SACA (ablation) | MCIE-E1 · CIE-Bench | IC: 0.7932 -> 0.8046 | Improved |
| add SACA alone to baseline (no BCCA) (ablation) | MCIE-E1 · CIE-Bench | L1: 0.0710 -> 0.0886 | Hurt |
| use larger / proprietary MLLM for instruction decomposition (3B to GPT-4o) (ablation) | MCIE-E1 · CIE-Bench | IC: 0.7403 -> 0.7873 | Improved |
| use MCIE-E1 instead of GoT | complex instruction editing · CIE-Bench | IC: 0.6491 -> 0.8046 | Improved |
| use MCIE-E1 (two-stage) instead of GoT on simple edits | MagicBrush · MagicBrush (simple instructions) | L1: 0.0645 -> 0.0669 | Mixed |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- AnyEdit — AnyEdit: Mastering Unified High-Quality Image Editing for Any Idea (CVPR 2025) · better on 1 of 1 shared result
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
- InsightEdit: Towards Better Instruction Following for Image Editing (CVPR 2025 · similarity 0.92)
- Guiding Instruction-based Image Editing via Multimodal Large Language Models (ICLR 2024 · similarity 0.92)
- SmartEdit: Exploring Complex Instruction-based Image Editing with Multimodal Large Language Models (CVPR 2024 · similarity 0.92)
- Balancing Preservation and Modification: A Region and Semantic Aware Metric for Instruction-Based Image Editing (ICML 2025 · similarity 0.91)
- Text-Vision Co-Instructed Image Editing (NeurIPS 2026 · similarity 0.89)
- Enabling Instructional Image Editing with In-Context Generation in Large Scale Diffusion Transformer (NeurIPS 2025 · similarity 0.89)
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.
