One-Step Diffusion Policy: Fast Visuomotor Policies via Diffusion Distillation

• URL: http://arxiv.org/abs/2410.21257v1
• Date: Mon, 28 Oct 2024 17:54:31 GMT
• Title: One-Step Diffusion Policy: Fast Visuomotor Policies via Diffusion Distillation
• Authors: Zhendong Wang, Zhaoshuo Li, Ajay Mandlekar, Zhenjia Xu, Jiaojiao Fan, Yashraj Narang, Linxi Fan, Yuke Zhu, Yogesh Balaji, Mingyuan Zhou, Ming-Yu Liu, Yu Zeng,
• Abstract summary: OneStep Diffusion Policy (OneDP) is a novel approach that distills knowledge from pre-trained diffusion policies into a single-step action generator. OneDP significantly accelerates response times for robotic control tasks.
• Score: 80.71541671907426
• License:
• Abstract: Diffusion models, praised for their success in generative tasks, are increasingly being applied to robotics, demonstrating exceptional performance in behavior cloning. However, their slow generation process stemming from iterative denoising steps poses a challenge for real-time applications in resource-constrained robotics setups and dynamically changing environments. In this paper, we introduce the One-Step Diffusion Policy (OneDP), a novel approach that distills knowledge from pre-trained diffusion policies into a single-step action generator, significantly accelerating response times for robotic control tasks. We ensure the distilled generator closely aligns with the original policy distribution by minimizing the Kullback-Leibler (KL) divergence along the diffusion chain, requiring only $2\%$-$10\%$ additional pre-training cost for convergence. We evaluated OneDP on 6 challenging simulation tasks as well as 4 self-designed real-world tasks using the Franka robot. The results demonstrate that OneDP not only achieves state-of-the-art success rates but also delivers an order-of-magnitude improvement in inference speed, boosting action prediction frequency from 1.5 Hz to 62 Hz, establishing its potential for dynamic and computationally constrained robotic applications. We share the project page at https://research.nvidia.com/labs/dir/onedp/.

Related papers

• PIVOT-R: Primitive-Driven Waypoint-Aware World Model for Robotic Manipulation [68.17081518640934]
  We propose a PrIrmitive-driVen waypOinT-aware world model for Robotic manipulation (PIVOT-R) PIVOT-R consists of a Waypoint-aware World Model (WAWM) and a lightweight action prediction module. Our PIVOT-R outperforms state-of-the-art open-source models on the SeaWave benchmark, achieving an average relative improvement of 19.45% across four levels of instruction tasks.
  arXiv  Detail & Related papers  (2024-10-14T11:30:18Z)
• Autoregressive Action Sequence Learning for Robotic Manipulation [32.9580007141312]
  We design a simple yet effective autoregressive architecture for robotic manipulation tasks. We propose the Chunking Causal Transformer (CCT), which extends the next-single-token prediction of causal transformers to support multi-token prediction in a single pass. Based on CCT, we propose the Autoregressive Policy (ARP) model, which learns to generate action sequences autoregressively.
  arXiv  Detail & Related papers  (2024-10-04T04:07:15Z)
• Bayesian-Optimized One-Step Diffusion Model with Knowledge Distillation for Real-Time 3D Human Motion Prediction [2.402745776249116]
  We propose training a one-step multi-layer perceptron-based (MLP-based) diffusion model for motion prediction using knowledge distillation and Bayesian optimization. Our model can significantly improve the inference speed, achieving real-time prediction without noticeable degradation in performance.
  arXiv  Detail & Related papers  (2024-09-19T04:36:40Z)
• HiRT: Enhancing Robotic Control with Hierarchical Robot Transformers [12.373320641721344]
  Large Vision-Language-Action (VLA) models have shown promise in robotic control due to their impressive generalization ability. Their reliance on VLM backends with billions of parameters leads to high computational costs and latency inference. This paper proposes HiRT, a Hierarchical Robot Transformer framework that enables flexible frequency and performance trade-off.
  arXiv  Detail & Related papers  (2024-09-12T09:18:09Z)
• ManiCM: Real-time 3D Diffusion Policy via Consistency Model for Robotic Manipulation [16.272352213590313]
  Diffusion models have been verified to be effective in generating complex distributions from natural images to motion trajectories. Recent methods show impressive performance in 3D robotic manipulation tasks, whereas they suffer from severe runtime inefficiency due to multiple denoising steps. We propose a real-time robotic manipulation model named ManiCM that imposes the consistency constraint to the diffusion process.
  arXiv  Detail & Related papers  (2024-06-03T17:59:23Z)
• Score identity Distillation: Exponentially Fast Distillation of Pretrained Diffusion Models for One-Step Generation [61.03530321578825]
  We introduce Score identity Distillation (SiD), an innovative data-free method that distills the generative capabilities of pretrained diffusion models into a single-step generator. SiD not only facilitates an exponentially fast reduction in Fr'echet inception distance (FID) during distillation but also approaches or even exceeds the FID performance of the original teacher diffusion models.
  arXiv  Detail & Related papers  (2024-04-05T12:30:19Z)
• One-Step Diffusion Distillation via Deep Equilibrium Models [64.11782639697883]
  We introduce a simple yet effective means of distilling diffusion models directly from initial noise to the resulting image. Our method enables fully offline training with just noise/image pairs from the diffusion model. We demonstrate that the DEQ architecture is crucial to this capability, as GET matches a $5times$ larger ViT in terms of FID scores.
  arXiv  Detail & Related papers  (2023-12-12T07:28:40Z)
• Learning Bipedal Walking for Humanoids with Current Feedback [5.429166905724048]
  We present an approach for overcoming the sim2real gap issue for humanoid robots arising from inaccurate torque-tracking at the actuator level. Our approach successfully trains a unified, end-to-end policy in simulation that can be deployed on a real HRP-5P humanoid robot to achieve bipedal locomotion.
  arXiv  Detail & Related papers  (2023-03-07T08:16:46Z)
• Leveraging Sequentiality in Reinforcement Learning from a Single Demonstration [68.94506047556412]
  We propose to leverage a sequential bias to learn control policies for complex robotic tasks using a single demonstration. We show that DCIL-II can solve with unprecedented sample efficiency some challenging simulated tasks such as humanoid locomotion and stand-up.
  arXiv  Detail & Related papers  (2022-11-09T10:28:40Z)
• A Self-Paced Mixed Distillation Method for Non-Autoregressive Generation [135.84684279852098]
  Non-Autoregressive (NAR) models significantly under-perform Auto-regressive (AR) models on various language generation tasks. Among the NAR models, BANG is the first large-scale pre-training model on English un-labeled raw text corpus. We propose a novel self-paced mixed distillation method to further improve the generation quality of BANG.
  arXiv  Detail & Related papers  (2022-05-23T09:54:53Z)

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.