统一服装生成模型通过正交化设计，解决多任务训练时的干扰问题。
OrthoTryOn: Geometric Orthogonalization for Conflict-Free Unified Fashion Generation
- 在共享LoRA模块中引入任务特异性正交旋转，分离不同任务特征空间。
- 实验显示其性能超越独立训练的专用模型，且在多个基准上达到顶尖水平。
- 无需额外参数，适合希望统一部署多种服装生成任务的研究者。
统一服装生成将虚拟试穿与服装重建等任务整合到单一模型中，以降低特定任务适配成本。然而，跨语义差异任务的简单参数共享会引发严重任务间梯度冲突，导致负迁移。本文提出OrthoTryOn框架，在共享低秩适应（LoRA）模块中通过正交子空间投影（OSP）对瓶颈特征进行任务特异性正交旋转，将其映射至解耦坐标系。为缓解推理时残余语义耦合，进一步提出无参的费舍尔引导负向引导（FNG）策略，利用对角费舍尔信息量化任务间敏感性重叠，通过无分类器引导显式排斥最易混淆任务的生成轨迹。大量实验表明，OrthoTryOn有效避免了传统统一训练中的严重性能下降，甚至优于独立训练的任务专用模型，在多个基准上实现领先结果，并在多种扩散模型主干上表现出强泛化能力。代码已公开于https://github.com/NJU-PCALab/OrthoTryOn。
原文摘要 · Abstract (English)
Unified fashion generation integrates tasks like virtual try-on and garment reconstruction into a single model to reduce task-specific adaptation costs. However, naive parameter sharing across semantically distinct tasks induces negative transfer through severe inter-task gradient conflict. We propose OrthoTryOn, a unified framework mitigating this interference within a shared Low-Rank Adaptation (LoRA) module. Its Orthogonal Subspace Projection (OSP) applies task-specific orthogonal rotations to bottleneck features, mapping them into decorrelated coordinate frames. To address residual semantic coupling at inference time, we further propose Fisher-guided Negative Guidance (FNG), a parameter-free strategy that utilizes diagonal Fisher information to quantify inter-task sensitivity overlap and explicitly repels generation trajectories from the most confusable task via Classifier-Free Guidance. Extensive experiments demonstrate that OrthoTryOn avoids the severe performance degradation typical of naive unified training and even surpasses independently trained task-specific models, achieving state-of-the-art results across multiple benchmarks while generalizing robustly across diverse diffusion backbones. Code is available at https://github.com/NJU-PCALab/OrthoTryOn.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。