[论文解读] LTC-SE: Expanding the Potential of Liquid Time-Constant Neural Networks for Scalable AI and Embedded Systems

Michael Bidollahkhani, Ferhat Atasoy+1 more|arXiv (Cornell University)|Apr 18, 2023

Advanced Memory and Neural Computing被引用 7

一句话总结

LTC-SE 提供与 TensorFlow 2.x 兼容的增强型 LTC 神经网络库，具有更灵活的配置和更好的嵌入式系统代码组织，并在可用性方面优于先前的 LTC 实现。

ABSTRACT

We present LTC-SE, an improved version of the Liquid Time-Constant (LTC) neural network algorithm originally proposed by Hasani et al. in 2021. This algorithm unifies the Leaky-Integrate-and-Fire (LIF) spiking neural network model with Continuous-Time Recurrent Neural Networks (CTRNNs), Neural Ordinary Differential Equations (NODEs), and bespoke Gated Recurrent Units (GRUs). The enhancements in LTC-SE focus on augmenting flexibility, compatibility, and code organization, targeting the unique constraints of embedded systems with limited computational resources and strict performance requirements. The updated code serves as a consolidated class library compatible with TensorFlow 2.x, offering comprehensive configuration options for LTCCell, CTRNN, NODE, and CTGRU classes. We evaluate LTC-SE against its predecessors, showcasing the advantages of our optimizations in user experience, Keras function compatibility, and code clarity. These refinements expand the applicability of liquid neural networks in diverse machine learning tasks, such as robotics, causality analysis, and time-series prediction, and build on the foundational work of Hasani et al.

研究动机与目标

• 提高资源受限嵌入式系统中 LTC 基模型的灵活性和兼容性。
• 将 LTC 相关架构（LTCCell、CTRNN、NODE、CTGRU）整合为一个可用库。
• 改进代码组织和用户体验，促进在机器人技术、因果分析和时序任务中的应用。

提出的方法

• 将 LTC-SE 作为一个扩展的、与 TensorFlow 2.x 兼容的类库进行开发。
• 提供可配置的 LTCCell、CTRNN、NODE 和 CTGRU 实现。
• 提升代码清晰度、Keras 兼容性和整体可用性。
• 在与先前的 LTC 实现比较时，重点评估用户体验和可读性。

实验结果

研究问题

• RQ1 相比于先前的 LTC 实现，LTC-SE 是否提升了用户体验和 Keras 兼容性？
• RQ2 LTC-SE 是否扩展了液态神经网络在机器人、因果分析和时序预测等任务的实际适用性？
• RQ3 LTC-SE 如何影响资源有限的嵌入式系统中的代码组织与可配置性？

主要发现

• LTC-SE 相对于前代产品在用户体验方面有所提升。
• LTC-SE 提供更好的 Keras 函数兼容性和更清晰的代码结构。
• 这些改进拓宽了液态神经网络在机器人、因果分析和时序任务中的适用性。