Research map / Image generation / Autoregressive and latent generation
EasyControl: Adding Efficient and Flexible Control for Diffusion Transformer
Yuxuan Zhang, Yirui Yuan, Yiren Song, Haofan Wang, Jiaming Liu
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 KV cache for condition features (ablation) | EasyControl · Single-condition inference time | time (s): 38.9 -> 25.0 (-36%) | Improved |
| add Position-Aware Training Paradigm (ablation) | EasyControl · Single-condition inference time | time (s): 22.4 -> 16.3 | Improved |
| add KV cache and PATP together (ablation) | EasyControl · Dual-condition inference time | time (s): 72.4 -> 18.3 | Improved |
| add KV cache (ablation) | EasyControl · Dual-condition inference time | time (s): 36.7 -> 18.3 | Improved |
| add Position-Aware Training Paradigm (ablation) | EasyControl · Dual-condition inference time | time (s): 29.9 -> 18.3 | Improved |
| add Condition Injection LoRA instead of standard LoRA (ablation) | EasyControl · multi-condition control | controllability | Improved |
| add Position-Aware Training Paradigm (ablation) | EasyControl · high-resolution or non-square generation | image quality | Improved |
| add Causal Mutual Attention (ablation) | EasyControl · multi-condition control | control accuracy | Improved |
| replace ControlNet with EasyControl single-condition | DiT control · single-condition parameters | params: 3B -> 15M | Improved |
| replace ControlNet+IPA with EasyControl | DiT control · dual-condition parameters | params: 4B -> 30M | Improved |
| replace OminiControl with EasyControl | DiT control · COCO Canny | FID: 17.237 -> 16.074 | Improved |
| replace OminiControl with EasyControl | DiT control · COCO Depth | FID: 18.536 -> 20.394 | Hurt |
Later papers that compared with this method
5 later accepted papers report EasyControl as a baseline in their results tables; 4 of them report a better number on most of the results they share with it.
- Inpaint-Anywhere: Zero-Shot Multi-Identity Inpainting with Efficient Diffusion Transformer (AAAI 2026 · better on 8 of 8 shared results)
- Any2RSI: Controllable Remote Sensing Text-to-Image Generation via Any Control and Enriched Description (AAAI 2026 · better on 7 of 7 shared results)
- EchoGen: Generating Visual Echoes in Any Scene via Feed-Forward Subject-Driven Auto-Regressive Model (ICLR 2026 · better on 4 of 5 shared results)
- HierEdit: Region-Aware Hierarchical Diffusion for Efficient High-Resolution Editing (CVPR 2026 · better on 2 of 2 shared results)
- Cross-ControlNet: Training-Free Fusion of Multiple Conditions for Text-to-Image Generation (ICLR 2026 · better on 0 of 1 shared result)
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.
- OminiControl — OminiControl: Minimal and Universal Control for Diffusion Transformer (ICCV 2025) · better on 6 of 8 shared results
- ControlNet — ControlNet++: Improving Conditional Controls with Efficient Consistency Feedback (ECCV 2024) · better on 3 of 4 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
- Dynamic Diffusion Transformer (ICLR 2025 · similarity 0.89)
- Simplifying Control Mechanism in Text-to-Image Diffusion Models (AAAI 2025 · similarity 0.88)
- FlexControl: Computation-Aware Conditional Control with Differentiable Router for Text-to-Image Generation (ICML 2025 · similarity 0.88)
- FlexiDiT: Your Diffusion Transformer Can Easily Generate High-Quality Samples with Less Compute (CVPR 2025 · similarity 0.88)
- A Simple Approach to Unifying Diffusion-based Conditional Generation (ICLR 2025 · similarity 0.88)
- UniDB: A Unified Diffusion Bridge Framework via Stochastic Optimal Control (ICML 2025 · 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.
