Research map / Video / Video generation
Composing Concepts from Images and Videos via Concept-prompt Binding
Xianghao Kong, Zeyu Zhang, Yuwei Guo, Zhuoran Zhao, Songchun Zhang, Anyi Rao
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 hierarchical binder (global plus per-block) (ablation) | global binder baseline · Concept composition ablation (user study) | Overall Quality: 2.34 -> 2.81 | Improved |
| add prompt diversification (ablation) | hierarchical binder · Concept composition ablation (user study) | Overall Quality: 2.81 -> 3.26 | Improved |
| add absorbent token (ablation) | Hrc+Div · Concept composition ablation (user study) | Overall Quality: 3.26 -> 3.47 | Improved |
| add temporal disentanglement strategy (ablation) | Hrc+Div · Concept composition ablation (user study) | Overall Quality: 3.26 -> 3.82 | Improved |
| add temporal disentanglement strategy (ablation) | Hrc+Div+Abs · Concept composition ablation (user study) | Overall Quality: 3.47 -> 4.40 | Improved |
| add absorbent token (ablation) | Hrc+Div+TDS · Concept composition ablation (user study) | Overall Quality: 3.82 -> 4.40 | Improved |
| add two-stage inverted training (ablation) | BiCo · Concept composition ablation (user study) | Overall Quality: 2.58 -> 4.40 | Improved |
| add two-stage training without inversion (ablation) | BiCo · Concept composition ablation (user study) | Overall Quality: 2.58 -> 3.61 | Improved |
| add inversion toward high noise levels in first stage (ablation) | two-stage BiCo · Concept composition ablation (user study) | Overall Quality: 3.61 -> 4.40 | Improved |
| full BiCo vs DualReal on human overall quality | DualReal · Concept composition (40 test cases) | Overall Quality: 3.00 -> 4.64 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- DualReal — DualReal: Adaptive Joint Training for Lossless Identity-Motion Fusion in Video Customization (ICCV 2025) · better on 6 of 6 shared results
- DreamVideo — DreamVideo: Composing Your Dream Videos with Customized Subject and Motion (CVPR 2024) · better on 6 of 6 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
- ConceptBed: Evaluating Concept Learning Abilities of Text-to-Image Diffusion Models (AAAI 2024 · similarity 0.88)
- Non-confusing Generation of Customized Concepts in Diffusion Models (ICML 2024 · similarity 0.86)
- Imagine Before You Draw: Visual Prompt Engineering for Image Generation (NeurIPS 2026 · similarity 0.86)
- MC^2: Multi-concept Guidance for Customized Multi-concept Generation (CVPR 2025 · similarity 0.86)
- Learning to Infer Generative Template Programs for Visual Concepts (ICML 2024 · similarity 0.86)
- FreeCustom: Tuning-Free Customized Image Generation for Multi-Concept Composition (CVPR 2024 · similarity 0.85)
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.
