Diagnose-to-Generate: Eliminating Structural Distortions in Flow Models via Patch-Level Grounded Alignment

Flow-based image generators achieve strong semantic synthesis but still produce localized structural distortions, such as malformed hands and limbs. Existing post-training methods typically optimize global scalar rewards, which provide weak spatial credit assignment and may alter well-formed regions while correcting local defects. We present $\mathit{\text{Diagnose-to-Generate}}\mathit{\text{(D2G)}}$, a grounded alignment framework that separates visual diagnosis from generative optimization. We construct $\mathbf{\text{Achilles-600K}}$, a region-annotated dataset for structural anomalies, and train $\mathbf{\text{ARGUS}}$ through supervised cold-start and Region-Aware Group Relative Policy Optimization to predict anomaly categories and bounding boxes. These diagnoses are converted into confidence-weighted spatial rewards for our $\mathbf{\text{Patch-level}}\mathbf{\text{Flow-GRPO}}$. We further identify that global stochastic exploration can shift semantic regions across group rollouts, making patch-wise relative advantages spatially inconsistent. Our $\mathbf{\text{Trajectory-Preserving}}\mathbf{\text{Stochastic}}\mathbf{\text{Sampling}}\mathbf{\text{(TPSS)}}$ anchors early denoising with deterministic ODE trajectories and introduces stochasticity only at later steps, thereby limiting intra-group spatial drift while retaining local exploration. Experiments on FLUX.1-dev and SD3.5 show improved structural fidelity with preserved FID and CLIP scores. Consistent gains under HumanRefiner, zero-shot GPT-4.1, and blinded human evaluation further demonstrate that the improvements extend beyond the training reward model.