Quantization-Aware Imitation-Learning for Resource-Efficient Robotic Control
This enables more efficient deployment of imitation-learning models on resource-limited devices like robots and autonomous vehicles, though it is incremental as it builds on existing quantization techniques.
The paper tackles the computational cost of deep neural network-based policy models for robotic control by proposing a quantization framework that fine-tunes parameters to maintain accuracy under low-bit precision, achieving up to 3.7x speedup and 3.1x energy savings in evaluations on real hardware.
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.