paper-with-me

홈 › Papers

Reward Guided Latent Consistency Distillation

2024-03-16 · Jiachen Li, Weixi Feng, Wenhu Chen, William Yang Wang

Latent Consistency Distillation (LCD) has emerged as a promising paradigm for efficient text-to-image synthesis. By distilling a latent consistency model (LCM) from a pre-trained teacher latent diffusion model (LDM), LCD facilitates the generation of high-fidelity images within merely 2 to 4 inference steps. However, the LCM's efficient inference is obtained at the cost of the sample quality. In this paper, we propose compensating the quality loss by aligning LCM's output with human preference during training. Specifically, we introduce Reward Guided LCD (RG-LCD), which integrates feedback from a reward model (RM) into the LCD process by augmenting the original LCD loss with the objective of maximizing the reward associated with LCM's single-step generation. As validated through human evaluation, when trained with the feedback of a good RM, the 2-step generations from our RG-LCM are favored by humans over the 50-step DDIM samples from the teacher LDM, representing a 25-time inference acceleration without quality loss. As directly optimizing towards differentiable RMs can suffer from over-optimization, we take the initial step to overcome this difficulty by proposing the use of a latent proxy RM (LRM). This novel component serves as an intermediary, connecting our LCM with the RM. Empirically, we demonstrate that incorporating the LRM into our RG-LCD successfully avoids high-frequency noise in the generated images, contributing to both improved Fr\'echet Inception Distance (FID) on MS-COCO and a higher HPSv2.1 score on HPSv2's test set, surpassing those achieved by the baseline LCM.

📄 PDF Abstract BibTeX arXiv:2403.11027

Code (0)

등록된 구현이 없습니다.

Tasks

Image Generation

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…
Latent Diffusion Model Diffusion models applied to latent spaces, which are normally built with (Variational) Autoencoders.

Similar Papers 제목 키워드 기반

Trajectory Consistency Distillation: Improved Latent Consistency Distillation by Semi-Linear Consistency Function with Trajectory Mapping

2024-02-29 · Jianbin Zheng, Minghui Hu, Zhongyi Fan, Chaoyue Wang 외

Latent Consistency Model (LCM) extends the Consistency Model to the latent space and leverages the guided consistency distillation technique to achieve impressive performance in accelerating text-to-image synthesis. Howe…

Image Generation

VGGRPO: Towards World-Consistent Video Generation with 4D Latent Reward

2026-03-27 · Zhaochong An, Orest Kupyn, Théo Uscidda, Andrea Colaco 외 arxiv

Large-scale video diffusion models achieve impressive visual quality, yet often fail to preserve geometric consistency. Prior approaches improve consistency either by augmenting the generator with additional modules or a…

Video Generation

Latent Reward Registers for Diffusion Preference Alignment

2026-08-04 · Yuanshen Guan, Zipeng Feng, Chengru Song, Zhiwei Xiong 외 arxiv

Aligning diffusion models with human preferences usually relies on a sparse terminal reward evaluated on the final generated samples, which creates a severe temporal credit-assignment problem across the denoising process…

Reinforcement Learning

DOLLAR: Few-Step Video Generation via Distillation and Latent Reward Optimization

2024-12-20 · Zihan Ding, Chi Jin, Difan Liu, Haitian Zheng 외

Diffusion probabilistic models have shown significant progress in video generation; however, their computational efficiency is limited by the large number of sampling steps required. Reducing sampling steps often comprom…

Computational EfficiencyDiversityVideo Generation

DiLaDiff: Distilled Latent-Augmented Diffusion for Language Modeling

2026-05-22 · Jean-Marie Lemercier, Tomas Geffner, Karsten Kreis, Morteza Mardani 외 arxiv

Diffusion language models intrinsically fail to capture correlations between decoded tokens, which leads to a harsh trade-off between sampling quality and throughput. To solve this issue, we propose DiLaDiff, a variant o…