paper-with-me

Papers

Aligning Generative Denoising with Discriminative Objectives Unleashes Diffusion for Visual Perception

2025-04-15 · Ziqi Pang, Xin Xu, Yu-Xiong Wang

With the success of image generation, generative diffusion models are increasingly adopted for discriminative tasks, as pixel generation provides a unified perception interface. However, directly repurposing the generative denoising process for discriminative objectives reveals critical gaps rarely addressed previously. Generative models tolerate intermediate sampling errors if the final distribution remains plausible, but discriminative tasks require rigorous accuracy throughout, as evidenced in challenging multi-modal tasks like referring image segmentation. Motivated by this gap, we analyze and enhance alignment between generative diffusion processes and perception tasks, focusing on how perception quality evolves during denoising. We find: (1) earlier denoising steps contribute disproportionately to perception quality, prompting us to propose tailored learning objectives reflecting varying timestep contributions; (2) later denoising steps show unexpected perception degradation, highlighting sensitivity to training-denoising distribution shifts, addressed by our diffusion-tailored data augmentation; and (3) generative processes uniquely enable interactivity, serving as controllable user interfaces adaptable to correctional prompts in multi-round interactions. Our insights significantly improve diffusion-based perception models without architectural changes, achieving state-of-the-art performance on depth estimation, referring image segmentation, and generalist perception tasks. Code available at https://github.com/ziqipang/ADDP.

📄 PDF Abstract BibTeX arXiv:2504.11457

Code (1)

ziqipang/addp 공식 구현 pytorch

Tasks

Data AugmentationDenoisingDepth EstimationImage GenerationImage SegmentationSemantic Segmentation

Methods 이 논문이 사용한 방법론

Diffusion Diffusion models generate samples by gradually removing noise from a signal, and their training objective can be expressed as a reweighted variational lower-bound…

Similar Papers 제목 키워드 기반

From Noise to Order: Learning to Rank via Denoising Diffusion

2026-02-12 · Sajad Ebrahimi, Bhaskar Mitra, Negar Arabzadeh, Ye Yuan 외 arxiv

In information retrieval (IR), learning-to-rank (LTR) methods have traditionally limited themselves to discriminative machine learning approaches that model the probability of the document being relevant to the query giv…

Information Retrieval

Probing Diffusion Denoising Dynamics for Contrastive Representation Learning

2026-07-10 · Yasong Dai, Zeeshan Hayder, David Ahmedt-Aristizabal, Hongdong Li arxiv

Text-to-image diffusion models exhibit unprecedented generative capability and contain rich intermediate representations that can be useful for discriminative vision tasks. Motivated by this observation, we study a focus…

Representation LearningContrastive Learning

Self-Supervised Representation-Guided Generative Dataset Distillation

2026-08-04 · Mingzhuo Li, Guang Li, Linfeng Ye, Jiafeng Mao 외 arxiv

Dataset distillation compresses a large training set into a compact synthetic set while retaining its downstream utility. Most existing methods target randomly initialized networks, whereas modern vision systems often ad…

Step-level Denoising-time Diffusion Alignment with Multiple Objectives

2026-04-15 · Qi Zhang, Dawei Wang, Shaofeng Zou arxiv

Reinforcement learning (RL) has emerged as a powerful tool for aligning diffusion models with human preferences, typically by optimizing a single reward function under a KL regularization constraint. In practice, however…

Reinforcement Learning

Improving Visual Representation Alignment Generation with GRPO

2026-05-30 · Shentong Mo, Sukmin Yun arxiv

Recent diffusion transformers have demonstrated strong image synthesis capabilities but remain inefficient to train due to weak alignment between generative and discriminative representations. While representation alignm…