# From Corpora to Co-Evolving Capabilities: Capability-Centric Data Design for Generalist Image Generation

arXiv: https://arxiv.org/abs/2608.18076 · HTML: https://mlsanity.com/paper/2608.18076/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 41 of its 45 references that are in the map.
- First posted on arXiv 18 Aug 2026.

## Builds on (most cited first, 30 of 41)

- [LAION-5B: An open large-scale dataset for training next generation image-text models](https://mlsanity.com/paper/2210.08402/) (2022)
- [Wan: Open and Advanced Large-Scale Video Generative Models](https://mlsanity.com/paper/2503.20314/) (2025)
- [DINOv3](https://mlsanity.com/paper/2508.10104/) (2025)
- [Qwen-Image Technical Report](https://mlsanity.com/paper/2508.02324/) (2025)
- [SAM 3: Segment Anything with Concepts](https://mlsanity.com/paper/2511.16719/) (2025)
- [Emerging Properties in Unified Multimodal Pretraining](https://mlsanity.com/paper/2505.14683/) (2025)
- [DataComp: In search of the next generation of multimodal datasets](https://mlsanity.com/paper/2304.14108/) (2023)
- [DoReMi: Optimizing Data Mixtures Speeds Up Language Model Pretraining](https://mlsanity.com/paper/2305.10429/) (2023)
- [OmniGen: Unified Image Generation](https://mlsanity.com/paper/2409.11340/) (2024)
- [The Faiss library](https://mlsanity.com/paper/2401.08281/) (2024)
- [Z-Image: An Efficient Image Generation Foundation Model with Single-Stream Diffusion Transformer](https://mlsanity.com/paper/2511.22699/) (2025)
- [Seedream 4.0: Toward Next-generation Multimodal Image Generation](https://mlsanity.com/paper/2509.20427/) (2025)
- [AnyEdit: Mastering Unified High-Quality Image Editing for Any Idea](https://mlsanity.com/paper/2411.15738/) (2024)
- [Multimodal C4: An Open, Billion-scale Corpus of Images Interleaved with Text](https://mlsanity.com/paper/2304.06939/) (2023)
- [HQ-Edit: A High-Quality Dataset for Instruction-based Image Editing](https://mlsanity.com/paper/2404.09990/) (2024)
- [AnyText: Multilingual Visual Text Generation And Editing](https://mlsanity.com/paper/2311.03054/) (2023)
- [OmniEdit: Building Image Editing Generalist Models Through Specialist Supervision](https://mlsanity.com/paper/2411.07199/) (2024)
- [LongCat-Image Technical Report](https://mlsanity.com/paper/2512.07584/) (2025)
- [ShareGPT-4o-Image: Aligning Multimodal Models with GPT-4o-Level Image Generation](https://mlsanity.com/paper/2506.18095/) (2025)
- [What If We Recaption Billions of Web Images with LLaMA-3?](https://mlsanity.com/paper/2406.08478/) (2024)
- [Zero-shot Image Editing with Reference Imitation](https://mlsanity.com/paper/2406.07547/) (2024)
- [Pico-Banana-400K: A Large-Scale Dataset for Text-Guided Image Editing](https://mlsanity.com/paper/2510.19808/) (2025)
- [LivePhoto: Real Image Animation with Text-guided Motion Control](https://mlsanity.com/paper/2312.02928/) (2023)
- [Wear-Any-Way: Manipulable Virtual Try-on via Sparse Correspondence Alignment](https://mlsanity.com/paper/2403.12965/) (2024)
- [On the Scalability of Diffusion-based Text-to-Image Generation](https://mlsanity.com/paper/2404.02883/) (2024)
- [FireRed-Image-Edit-1.0 Technical Report](https://mlsanity.com/paper/2602.13344/) (2026)
- [Qwen-Image-2.0 Technical Report](https://mlsanity.com/paper/2605.10730/) (2026)
- [Tunnel Try-on: Excavating Spatial-temporal Tunnels for High-quality Virtual Try-on in Videos](https://mlsanity.com/paper/2404.17571/) (2024)
- [DreamOmni: Unified Image Generation and Editing](https://mlsanity.com/paper/2412.17098/) (2024)
- [BizGen: Advancing Article-level Visual Text Rendering for Infographics Generation](https://mlsanity.com/paper/2503.20672/) (2025)

## Built on by (most cited first, 2 of 2)

- [Exploring the Performance Frontier of Compact Unified Image Generation Models](https://mlsanity.com/paper/2608.20334/) (2026)
- [CommerceGen-Studio: An End-to-End Framework for Commercial-Grade Main-Image Generation](https://mlsanity.com/paper/2610.20603/) (2026)

## Abstract (arXiv)

Large-scale image generation has benefited from advances in data scale, quality, rebalancing, and recaptioning, yet conventional pipelines typically optimize task-specific datasets in isolation. A central challenge is not only how to curate each task-specific corpus, but also how to organize heterogeneous supervision according to the dependencies among generative capabilities. We present a capability-driven data infrastructure that couples capability-specific supervision construction with capability-aligned curriculum scheduling. Its three specialized yet interoperable data engines build complementary relational supervision for text-image grounding, inter-image transformation, and image-knowledge association, while caption experts align T2I and editing supervision across tasks and granularities. A multi-stage curriculum jointly evolves task composition, visual-concept distribution, data quality, and image resolution along the dependency order of capability acquisition, with capability-aware evaluation closing the loop through targeted retrieval, expert construction, and gap-aware resampling. At scale, the framework curates a 440M-image T2I corpus, 120M editing pairs, and over 27M image-entity pairs. With this infrastructure, we train multimodal diffusion models at two scales from scratch, with 3B and 6B sizes respectively. We conduct quantitative evaluation on CPI-Bench, along with qualitative evaluations across diverse text-to-image and editing scenarios. Experimental results present broad visual coverage, versatile rendering, and effective transfer across generative capabilities.