Diffusion Distillation With Direct Preference Optimization For Efficient 3D LiDAR Scene Completion

• URL: http://arxiv.org/abs/2504.11447v2
• Date: Wed, 16 Apr 2025 02:02:05 GMT
• Title: Diffusion Distillation With Direct Preference Optimization For Efficient 3D LiDAR Scene Completion
• Authors: An Zhao, Shengyuan Zhang, Ling Yang, Zejian Li, Jiale Wu, Haoran Xu, AnYang Wei, Perry Pengyun GU, Lingyun Sun,
• Abstract summary: This paper proposes Distillation-DPO, a novel diffusion distillation framework for LiDAR scene completion with preference aligment.<n>Our method is the first to explore adopting preference learning in distillation to the best of our knowledge and provide insights into preference-aligned distillation.
• Score: 25.55163699029964
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: The application of diffusion models in 3D LiDAR scene completion is limited due to diffusion's slow sampling speed. Score distillation accelerates diffusion sampling but with performance degradation, while post-training with direct policy optimization (DPO) boosts performance using preference data. This paper proposes Distillation-DPO, a novel diffusion distillation framework for LiDAR scene completion with preference aligment. First, the student model generates paired completion scenes with different initial noises. Second, using LiDAR scene evaluation metrics as preference, we construct winning and losing sample pairs. Such construction is reasonable, since most LiDAR scene metrics are informative but non-differentiable to be optimized directly. Third, Distillation-DPO optimizes the student model by exploiting the difference in score functions between the teacher and student models on the paired completion scenes. Such procedure is repeated until convergence. Extensive experiments demonstrate that, compared to state-of-the-art LiDAR scene completion diffusion models, Distillation-DPO achieves higher-quality scene completion while accelerating the completion speed by more than 5-fold. Our method is the first to explore adopting preference learning in distillation to the best of our knowledge and provide insights into preference-aligned distillation. Our code is public available on https://github.com/happyw1nd/DistillationDPO.

Related papers

• Denoising Score Distillation: From Noisy Diffusion Pretraining to One-Step High-Quality Generation [82.39763984380625]
  We introduce denoising score distillation (DSD), a surprisingly effective and novel approach for training high-quality generative models from low-quality data.<n>DSD pretrains a diffusion model exclusively on noisy, corrupted samples and then distills it into a one-step generator capable of producing refined, clean outputs.
  arXiv  Detail & Related papers  (2025-03-10T17:44:46Z)
• Inference-Time Diffusion Model Distillation [59.350789627086456]
  We introduce Distillation++, a novel inference-time distillation framework.<n>Inspired by recent advances in conditional sampling, our approach recasts student model sampling as a proximal optimization problem.<n>We integrate distillation optimization during reverse sampling, which can be viewed as teacher guidance.
  arXiv  Detail & Related papers  (2024-12-12T02:07:17Z)
• Distilling Diffusion Models to Efficient 3D LiDAR Scene Completion [25.517559974601813]
  Diffusion models have been applied to 3D LiDAR scene completion due to their strong training stability and high completion quality.<n>This paper proposes a novel distillation method tailored for 3D LiDAR scene completion models, dubbed $textbfScoreLiDAR$, which achieves efficient yet high-quality scene completion.
  arXiv  Detail & Related papers  (2024-12-04T17:57:25Z)
• Tuning Timestep-Distilled Diffusion Model Using Pairwise Sample Optimization [97.35427957922714]
  We present an algorithm named pairwise sample optimization (PSO), which enables the direct fine-tuning of an arbitrary timestep-distilled diffusion model.<n>PSO introduces additional reference images sampled from the current time-step distilled model, and increases the relative likelihood margin between the training images and reference images.<n>We show that PSO can directly adapt distilled models to human-preferred generation with both offline and online-generated pairwise preference image data.
  arXiv  Detail & Related papers  (2024-10-04T07:05:16Z)
• Target-Driven Distillation: Consistency Distillation with Target Timestep Selection and Decoupled Guidance [17.826285840875556]
  We introduce Target-Driven Distillation (TDD) to accelerate generative tasks of diffusion models. TDD adopts delicate selection strategy of target timesteps, increasing the training efficiency. It can be equipped with non-equidistant sampling and x0 clipping, enabling a more flexible and accurate way for image sampling.
  arXiv  Detail & Related papers  (2024-09-02T16:01:38Z)
• One Step Diffusion-based Super-Resolution with Time-Aware Distillation [60.262651082672235]
  Diffusion-based image super-resolution (SR) methods have shown promise in reconstructing high-resolution images with fine details from low-resolution counterparts. Recent techniques have been devised to enhance the sampling efficiency of diffusion-based SR models via knowledge distillation. We propose a time-aware diffusion distillation method, named TAD-SR, to accomplish effective and efficient image super-resolution.
  arXiv  Detail & Related papers  (2024-08-14T11:47:22Z)
• Scaling Diffusion Models to Real-World 3D LiDAR Scene Completion [25.69896680908217]
  3D LiDAR sensors are commonly used to collect sparse 3D point clouds from the scene. We propose extending diffusion models as generative models for images to achieve scene completion from a single 3D LiDAR scan. Our method can complete the scene given a single LiDAR scan as input, producing a scene with more details compared to state-of-the-art scene completion methods.
  arXiv  Detail & Related papers  (2024-03-20T10:19:05Z)
• Fast High-Resolution Image Synthesis with Latent Adversarial Diffusion Distillation [24.236841051249243]
  Distillation methods aim to shift the model from many-shot to single-step inference. We introduce Latent Adversarial Diffusion Distillation (LADD), a novel distillation approach overcoming the limitations of ADD. In contrast to pixel-based ADD, LADD utilizes generative features from pretrained latent diffusion models.
  arXiv  Detail & Related papers  (2024-03-18T17:51:43Z)
• BOOT: Data-free Distillation of Denoising Diffusion Models with Bootstrapping [64.54271680071373]
  Diffusion models have demonstrated excellent potential for generating diverse images. Knowledge distillation has been recently proposed as a remedy that can reduce the number of inference steps to one or a few. We present a novel technique called BOOT, that overcomes limitations with an efficient data-free distillation algorithm.
  arXiv  Detail & Related papers  (2023-06-08T20:30:55Z)
• DiffTAD: Temporal Action Detection with Proposal Denoising Diffusion [137.8749239614528]
  We propose a new formulation of temporal action detection (TAD) with denoising diffusion, DiffTAD. Taking as input random temporal proposals, it can yield action proposals accurately given an untrimmed long video.
  arXiv  Detail & Related papers  (2023-03-27T00:40:52Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.