DeCoRL is a reinforcement‑learning framework designed to enhance multi‑image composition (MICo) by decomposing overall consistency into object‑level consistency. It first establishes instruction‑relevant object correspondences across reference images and localizes these regions in the generated image. A VLM‑based reward model, trained on single‑object consistency annotations, provides fine‑grained, object‑wise feedback that is aggregated into a multi‑object consistency reward to guide RL post‑training. The authors also introduce DeCo‑Bench, a benchmark covering compositions from 2 to 6 reference images, and show consistent improvements on both DeCo‑Bench and the public MICo‑Bench.
Despite remarkable progress in text-guided image editing, generative models frequently fail to preserve visual object consistency, defined as the preservation of a subject's key attributes throughout the editing process. We address this limitation through three contributions.
PrismGPT is a Vision‑Language Model that generates structured, region‑aware photo‑editing plans from a single image, without relying on commercial black‑box tools. It learns to diagnose aesthetic issues globally and locally while predicting precise editing parameters, using proxy‑guided learning with operation decomposition and region‑aware aesthetic ranking to bootstrap the model. A competence‑based dynamic scheduler shifts training focus from proxy tasks to the main editing task as skills improve, and all reasoning traces for fine‑tuning are self‑synthesized by the model itself. Experiments on MIT‑Adobe FiveK and a new professionally retouched benchmark, SPIRE, show PrismGPT achieves state‑of‑the‑art results using only about 6% of the training data required by previous methods.
Recent breakthroughs in instruction-based image editing have captured significant attention, as models are now capable of handling real-world editing demands with the practicality required by everyday users. However, editing models trained primarily for single-turn edits often break down in multi-turn editing--the natural interactive setting where a user iteratively refines an image based on the model's own previous outputs.
arXiv:2608.22780v1 Announce Type: new Abstract: Reinforcement learning (RL) enables direct preference optimization for image editing through editing-specific rewards, which remain less developed due...
arXiv:2603. 03143v2 Announce Type: replace-cross Abstract: Leveraging the priors of 2D diffusion models for 3D editing has emerged as a promising paradigm.
RefVideo-6M is a new large-scale reference-guided editing dataset that includes 5 million video editing samples and 1 million image editing samples, each paired with about 6 million visual references. The dataset is constructed to avoid artifacts by using real, artifact‑free videos as targets and filtering input conditions with multiple editing experts, thereby providing reliable supervision. It enables models to learn fine‑grained visual correspondence beyond text‑only instructions and supports the training of a reference‑guided video editing model, Ref‑MoT, which shows improved visual quality, controllability, and reference consistency.
The paper introduces WeGenEditBench, a benchmark with 2,740 evaluation instances and 148 task types for image editing, focusing on local failure detection and prioritization for model improvement. It also presents WeGenEvoStudio, a multi‑agent harness that supports recursive self‑improvement by diagnosing failures, constructing retrieval‑grounded data, and adapting models. Experiments show that models such as GPT‑Image‑2 and Seedream‑5.0‑Pro excel in single‑image and text‑intensive editing, and evolution experiments demonstrate significant targeted gains on selected benchmarks.
arXiv:2610.16224v1 Announce Type: cross Abstract: Text-to-image models have made remarkable progress in visual fidelity and instruction following, yet complex compositional generation remains challen...
arXiv:2509. 24900v2 Announce Type: replace-cross Abstract: The performance of unified multimodal models for image generation and editing is fundamentally constrained by the quality and comprehensiveness of their training data.
SemantiScene-Edit is a video editing system that uses a multimodal encoder and a pretrained video diffusion transformer to follow natural‑language instructions while preserving the source video’s appearance and temporal structure. The encoder processes source frames and the instruction together, sending semantic queries to the generator via a trainable connector and cross‑attention. A dual‑path visual route concatenates source‑video latents with noisy target latents to provide dense spatial and temporal information, and the model is trained with flow matching and a three‑stage curriculum on over 1.2 million instruction‑conditioned records, with a 100‑case benchmark for evaluation.
ReasonUni introduces a reasoning-augmented reconstructive post‑training framework for unified multimodal models (UMMs) that couples image‑to‑text understanding and text‑to‑image generation. The framework first reasons about the desired edit and its target state, then generates the target image conditioned on this reasoning output. A tree‑structured reinforcement learning method, TreeGRPO, jointly optimizes the reasoning and generation modules, yielding consistent performance gains across multiple benchmarks.
arXiv:2605. 09233v2 Announce Type: replace-cross Abstract: Recent advances in visual generative models have enabled high-fidelity image editing guided by human instructions.