Research map / Image generation / Text-to-image diffusion
EchoGen: Cycle-Consistent Learning for Unified Layout-Image Generation and Understanding
Kai Zou, Hongbo Liu, Dian Zheng, Jianxiong Gao, Zhiwei Zhao, Bin Liu
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 Stage 2 Dual Joint Optimization after Stage 1 (ablation) | EchoGen · MS COCO | AP: 47.26 -> 52.38 | Improved |
| add Stage 3 Cycled RL after Stages 1+2 (ablation) | EchoGen · MS COCO | AP: 52.38 -> 54.61 | Improved |
| add Stage 3 RL directly to Stage 1 without Stage 2 (ablation) | EchoGen · MS COCO | AP: 47.26 -> 48.01 | No clear effect |
| add Stage 2 to Stage 1 (ablation) | EchoGen · MS COCO | FID: 21.98 -> 22.23 | No clear effect |
| add Stage 3 RL to Stages 1+2 (ablation) | EchoGen · MS COCO | FID: 22.23 -> 20.12 | Improved |
| add Stage 3 RL to Stage 1 only (ablation) | EchoGen · MS COCO | FID: 21.98 -> 20.32 | Improved |
| use random layouts instead of real-dataset layouts for RL (ablation) | EchoGen · MS COCO | AP: 54.61 -> 53.96 | No clear effect |
| use GPT-4o generated layouts instead of real-dataset layouts for RL (ablation) | EchoGen · MS COCO | AP: 54.61 -> 54.51 | No clear effect |
| enter Stage 2 with low base AP (below 20) (ablation) | EchoGen · MS COCO | AP | Hurt |
| enter Stage 3 RL with weak base accuracy (AP below 40) (ablation) | EchoGen · MS COCO | AP | Hurt |
| unified generation and grounding model (EchoGen) versus unified PlanGen | EchoGen · MS COCO | AP: 51.39 -> 54.61 | Improved |
| unified EchoGen versus best grounding-only CogVLM-grounding | EchoGen · Ref-L4 | Acc0.75: 70.77 -> 75.42 | 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.
- PlanGen — PlanGen: Towards Unified Layout Planning and Image Generation in Auto-Regressive Vision Language Models (ICCV 2025) · better on 8 of 9 shared results
- InstanceDiffusion — InstanceDiffusion: Instance-level Control for Image Generation (CVPR 2024) · 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
- Rethinking The Training And Evaluation of Rich-Context Layout-to-Image Generation (NeurIPS 2024 · similarity 0.88)
- InstanceAssemble: Layout-Aware Image Generation via Instance Assembling Attention (NeurIPS 2025 · similarity 0.88)
- SpotActor: Training-Free Layout-Controlled Consistent Image Generation (AAAI 2025 · similarity 0.88)
- PLACE: Adaptive Layout-Semantic Fusion for Semantic Image Synthesis (CVPR 2024 · similarity 0.87)
- Think, Plan, Paint: Layout-Aware Reasoning for Controllable Image Generation in Unified Models (NeurIPS 2026 · similarity 0.87)
- Seeing is Improving: Visual Feedback for Iterative Text Layout Refinement (CVPR 2026 · similarity 0.87)
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.
