Pith. sign in

REVIEW 6 cited by

IterComp: Iterative Composition-Aware Feedback Learning from Model Gallery for Text-to-Image Generation

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2410.07171 v2 pith:B3W6U25V submitted 2024-10-09 cs.CV

IterComp: Iterative Composition-Aware Feedback Learning from Model Gallery for Text-to-Image Generation

Xinchen Zhang , Ling Yang , Guohao Li , Yaqi Cai , Jiake Xie , Yong Tang , Yujiu Yang , Mengdi Wang
show 1 more author
Bin Cui
This is my paper
classification cs.CV
keywords modelscompositionaldiffusiongenerationcomposition-awarefeedbackitercomplearning
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

Advanced diffusion models like RPG, Stable Diffusion 3 and FLUX have made notable strides in compositional text-to-image generation. However, these methods typically exhibit distinct strengths for compositional generation, with some excelling in handling attribute binding and others in spatial relationships. This disparity highlights the need for an approach that can leverage the complementary strengths of various models to comprehensively improve the composition capability. To this end, we introduce IterComp, a novel framework that aggregates composition-aware model preferences from multiple models and employs an iterative feedback learning approach to enhance compositional generation. Specifically, we curate a gallery of six powerful open-source diffusion models and evaluate their three key compositional metrics: attribute binding, spatial relationships, and non-spatial relationships. Based on these metrics, we develop a composition-aware model preference dataset comprising numerous image-rank pairs to train composition-aware reward models. Then, we propose an iterative feedback learning method to enhance compositionality in a closed-loop manner, enabling the progressive self-refinement of both the base diffusion model and reward models over multiple iterations. Theoretical proof demonstrates the effectiveness and extensive experiments show our significant superiority over previous SOTA methods (e.g., Omost and FLUX), particularly in multi-category object composition and complex semantic alignment. IterComp opens new research avenues in reward feedback learning for diffusion models and compositional generation. Code: https://github.com/YangLing0818/IterComp

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 6 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Search Beyond What Can Be Taught: Evolving the Knowledge Boundary in Agentic Visual Generation

    cs.CV 2026-07 accept novelty 7.0

    A generator-specific knowledge boundary is discoverable by co-training: DPO teaches what can be internalized, then rejection finetuning calibrates search to only what remains external.

  2. MetaPoint: Unlocking Precise Spatial Control in Agentic Visual Generation

    cs.CV 2026-06 unverdicted novelty 7.0

    MetaPoint represents 2D coordinates as special tokens in visual generative models to enable precise spatial control using existing positional encodings without architectural modifications.

  3. LeapAlign: Post-Training Flow Matching Models at Any Generation Step by Building Two-Step Trajectories

    cs.CV 2026-04 unverdicted novelty 7.0

    LeapAlign fine-tunes flow matching models by constructing two consecutive leaps that skip multiple ODE steps with randomized timesteps and consistency weighting, enabling stable updates at any generation step.

  4. Search Beyond What Can Be Taught: Evolving the Knowledge Boundary in Agentic Visual Generation

    cs.CV 2026-07 conditional novelty 6.0

    A co-training framework (DPO on the generator, then rejection finetuning on the reasoner) discovers a generator-specific knowledge boundary that enables selective search for world-knowledge-grounded image generation, ...

  5. Search Beyond What Can Be Taught: Evolving the Knowledge Boundary in Agentic Visual Generation

    cs.CV 2026-07 conditional novelty 6.0

    Selective search plus generator-reasoner co-training improves knowledge-grounded image generation, but the reported gains are scored by the same VLM judge used to train the system.

  6. MRT: Masked Region Transformer for Layered Image Generation and Editing at Scale

    cs.CV 2026-05 unverdicted novelty 6.0

    Presents MRT, a 20B-parameter masked region diffusion model unifying text-to-layers, image-to-layers, and layers-to-layers tasks with an overflow-aware canvas layer for complete editable outputs.