EdgeLDR: Quaternion Low-Displacement Rank Neural Networks for Edge-Efficient Deep Learning
This work addresses efficiency challenges for edge computing in deep learning, offering a practical solution with incremental improvements over existing methods.
The paper tackles the problem of deploying deep neural networks on edge devices by introducing EdgeLDR, a framework that combines quaternion neural networks with block-circulant parameter structures to reduce memory and compute costs, achieving significant compression with competitive accuracy on tasks like CIFAR-10/100 and hyperspectral image classification.
Deploying deep neural networks on edge devices is often limited by the memory traffic and compute cost of dense linear operators. While quaternion neural networks improve parameter efficiency by coupling multiple channels through Hamilton products, they typically retain unstructured dense weights; conversely, structured matrices enable fast computation but are usually applied in the real domain. This paper introduces EdgeLDR, a practical framework for quaternion block-circulant linear and convolutional layers that combines quaternion channel mixing with block-circulant parameter structure and enables FFT-based evaluation through the complex adjoint representation. We present reference implementations of EdgeLDR layers and compare FFT-based computation against a naive spatial-domain realization of quaternion circulant products. FFT evaluation yields large empirical speedups over the naive implementation and keeps latency stable as block size increases, making larger compression factors computationally viable. We further integrate EdgeLDR layers into compact CNN and Transformer backbones and evaluate accuracy-compression trade-offs on 32x32 RGB classification (CIFAR-10/100, SVHN) and hyperspectral image classification (Houston 2013, Pavia University), reporting parameter counts and CPU/GPU latency. The results show that EdgeLDR layers provide significant compression with competitive accuracy.