mlsanityData to 7 Oct 2026

Quantization-Aware Imitation-Learning for Resource-Efficient Robotic Control

arXiv 2412.01034 ↗First posted 2 Dec 2024cs.RO cs.CV cs.LGcited by 25

On the map: Vision-language-action models → Efficient VLA inference and adaptation

Abstract (arXiv)

Deep neural network (DNN)-based policy models like vision-language-action (VLA) models are transformative in automating complex decision-making across applications by interpreting multi-modal data. However, scaling these models greatly increases computational costs, which presents challenges in fields like robot manipulation and autonomous driving that require quick, accurate responses. To address the need for deployment on resource-limited hardware, we propose a new quantization framework for IL-based policy models that fine-tunes parameters to enhance robustness against low-bit precision errors during training, thereby maintaining efficiency and reliability under constrained conditions. Our evaluations with representative robot manipulation for 4-bit weight-quantization on a real edge GPU demonstrate that our framework achieves up to 2.5x speedup and 2.5x energy savings while preserving accuracy. For 4-bit weight and activation quantized self-driving models, the framework achieves up to 3.7x speedup and 3.1x energy saving on a low-end GPU. These results highlight the practical potential of deploying IL-based policy models on resource-constrained devices.

Key facts. Cited by 25 papers in the map (data through 7 Oct 2026). Placed in Efficient VLA inference and adaptation, one of 13 directions in Vision-language-action models. Builds on 34 of its 48 references that are in the map. First posted on arXiv 2 Dec 2024.

Builds on

34 of 48 references.

References · 34

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2023GPT-4 Technical Reportcited by 18,132
  3. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
  4. 2017Proximal Policy Optimization Algorithmscited by 14,441
  5. 2023Visual Instruction Tuningcited by 7,792
  6. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,764
  7. 2022Flamingo: a Visual Language Model for Few-Shot Learningcited by 4,337
  8. 2023QLoRA: Efficient Finetuning of Quantized LLMscited by 3,355
  9. 2023Sigmoid Loss for Language Image Pre-Trainingcited by 3,048
  10. 2023RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Controlcited by 2,993
  11. 2024OpenVLA: An Open-Source Vision-Language-Action Modelcited by 2,777
  12. 2017CARLA: An Open Urban Driving Simulatorcited by 2,431
22 more
  1. 2016Xception: Deep Learning with Depthwise Separable Convolutionscited by 2,320
  2. 2023PaLM-E: An Embodied Multimodal Language Modelcited by 2,037
  3. 2022RT-1: Robotics Transformer for Real-World Control at Scalecited by 2,007
  4. 2017Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inferencecited by 1,548
  5. 2020D4RL: Datasets for Deep Data-Driven Reinforcement Learningcited by 1,406
  6. 2023LIBERO: Benchmarking Knowledge Transfer for Lifelong Robot Learningcited by 1,338
  7. 2018DeepMind Control Suitecited by 935
  8. 2023Open X-Embodiment: Robotic Learning Datasets and RT-X Modelscited by 849
  9. 2019Learned Step Size Quantizationcited by 604
  10. 2016DoReFa-Net: Training Low Bitwidth Convolutional Neural Networks with Low Bitwidth Gradientscited by 557
  11. 2018Quantizing deep convolutional networks for efficient inference: A whitepapercited by 507
  12. 2018PACT: Parameterized Clipping Activation for Quantized Neural Networkscited by 456
  13. 2019Exploring the Limitations of Behavior Cloning for Autonomous Drivingcited by 343
  14. 2024TinyVLA: Towards Fast, Data-Efficient Vision-Language-Action Models for Robotic Manipulationcited by 331
  15. 2018Distributed Distributional Deterministic Policy Gradientscited by 195
  16. 2023LoftQ: LoRA-Fine-Tuning-Aware Quantization for Large Language Modelscited by 195
  17. 2021End-to-End Urban Driving by Imitating a Reinforcement Learning Coachcited by 191
  18. 2017Visualizing and Understanding Atari Agentscited by 141
  19. 2018CIRL: Controllable Imitative Reinforcement Learning for Vision-based Self-drivingcited by 96
  20. 2018NVIDIA Tensor Core Programmability, Performance & Precisioncited by 77
  21. 2019QuaRL: Quantization for Fast and Environmentally Sustainable Reinforcement Learningcited by 19
  22. 2018Low Precision Policy Distillation with Application to Low-Power, Real-time Sensation-Cognition-Action Loop with Neuromorphic Computingcited by 2

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 11; 2026: 14.

25 papers.

Sort:

By area:

  1. 2025PD-VLA: Accelerating Vision-Language-Action Model Integrated with Action Chunking via Parallel Decodingcited by 88
  2. 2025VLA-Cache: Efficient Vision-Language-Action Manipulation via Adaptive Token Cachingcited by 83
  3. 2025SP-VLA: A Joint Model Scheduling and Token Pruning Approach for VLA Model Accelerationcited by 56
  4. 2025SpecPrune-VLA: Accelerating Vision-Language-Action Models via Action-Aware Self-Speculative Pruningcited by 39
  5. 2025CEED-VLA: Consistency Vision-Language-Action Model with Early-Exit Decodingcited by 38
  6. 2025A Survey on Efficient Vision-Language-Action Modelscited by 34
  7. 2025Spec-VLA: Speculative Decoding for Vision-Language-Action Models with Relaxed Acceptancecited by 31
  8. 2025Think Twice, Act Once: Token-Aware Compression and Action Reuse for Efficient Inference in Vision-Language-Action Modelscited by 30
  9. 2026QVLA: Not All Channels Are Equal in Vision-Language-Action Model's Quantizationcited by 23
  10. 2025RLRC: Reinforcement Learning-based Recovery for Compressed Vision-Language-Action Modelscited by 20
  11. 2026FASTER: Rethinking Real-Time Flow VLAscited by 20
  12. 2025SQAP-VLA: A Synergistic Quantization-Aware Pruning Framework for High-Performance Vision-Language-Action Modelscited by 15
13 more
  1. 2025Bridging the Semantic-Action Gap in Visual Token Pruning for Efficient VLA Inferencecited by 14
  2. 2026KERV: Kinematic-Rectified Speculative Decoding for Embodied VLA Modelscited by 9
  3. 2026DySL-VLA: Efficient Vision-Language-Action Model Inference via Dynamic-Static Layer-Skipping for Robot Manipulationcited by 8
  4. 2026Shallow-π: Knowledge Distillation for Flow-based VLAscited by 7
  5. 2026ActQuant: Sub-4-bit Action-Guided Quantization for Vision-Language-Action Modelscited by 7
  6. 2026HeiSD: Hybrid Speculative Decoding for Embodied Vision-Language-Action Models with Kinematic Awarenesscited by 6
  7. 2026Efficient Long-Horizon Vision-Language-Action Models via Static-Dynamic Disentanglementcited by 5
  8. 2026OxyGen: Unified KV Cache Management for VLA Inference under Multi-Task Parallelismcited by 4
  9. 2026Jetson-PI: Towards Onboard Real-Time Robot Control via Foresight-Aligned Asynchronous Inferencecited by 3
  10. 2026From Inference Efficiency to Embodied Efficiency: Revisiting Efficiency Metrics for Vision-Language-Action Modelscited by 2
  11. 2026ElegantVLA: Learning When to Think for Efficient Vision-Language-Action Modelscited by 2
  12. 2026Learning to Accelerate Vision-Language-Action Models through Adaptive Visual Token Cachingcited by 1
  13. 2026Beyond Attention Magnitude: Leveraging Inter-layer Rank Consistency for Efficient Vision-Language-Action Modelscited by 1

Similar, not cited

Close in wording, no citation either way

  1. 2025Saliency-Aware Quantized Imitation Learning for Efficient Robotic Control
  2. 2026HBVLA: Pushing 1-Bit Post-Training Quantization for Vision-Language-Action Models
  3. 2026FoldQuantVLA: Native Low-Bit Quantization of Vision-Language-Action Models via Consistent Folding
  4. 2025BitVLA: 1-bit Vision-Language-Action Models for Robotics Manipulation
  5. 2025Towards Accessible Physical AI: LoRA-Based Fine-Tuning of VLA Models for Real-World Robot Control
  6. 2026A Motion-Aware Vector Quantization Framework with Centroid Reuse for Efficient VLA Inference
  7. 2026Characterizing Vision-Language-Action Models across XPUs: Constraints and Acceleration for On-Robot Deployment
  8. 2023Prompt, Plan, Perform: LLM-based Humanoid Control via Quantized Imitation Learning