A novel deep neural network (DNN) architecture is proposed wherein the filtering and linear transform are realized solely with product quantization (PQ). This results in a natural implementation via content addressable memory (CAM), which transcends regular DNN layer operations and requires only simple table lookup. Two schemes are developed for the end-to-end PQ prototype training, namely, through angle- and distance-based similarities, which differ in their multiplicative and additive natures with different complexity-accuracy tradeoffs. Even more, the distance-based scheme constitutes a truly multiplier-free DNN solution. Experiments confirm the feasibility of such Product-Quantized Content Addressable Memory Network (PECAN), which has strong implication on hardware-efficient deployments especially for in-memory computing.
@article{arxiv.2208.13571,
title = {PECAN: A Product-Quantized Content Addressable Memory Network},
author = {Jie Ran and Rui Lin and Jason Chun Lok Li and Jiajun Zhou and Ngai Wong},
journal= {arXiv preprint arXiv:2208.13571},
year = {2022}
}