CIM-NET: A Video Denoising Deep Neural Network Model Optimized for Computing-in-Memory Architectures
IEEE Signal Processing Letters · 2026 · 33 · 1591-1595
DOI: 10.1109/lsp.2026.3680452 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Signal Processing Letters, 2026. Verfasst von Shan Gao und 4 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 29 Literaturangaben. Gefördert von: China Mobile Communications Group Company.
Abstract (Originalsprache)
Deep neural network (DNN)-based video denoising has demonstrated significant performance, but deploying state-of-the-art models on edge devices remains challenging due to stringent real-time and energy efficiency requirements. Computing-in-Memory (CIM) chips offer a promising solution by integrating computation within memory cells, enabling rapid matrix-vector multiplication (MVM). However, existing DNN models are often designed without considering CIM architectural constraints, thus limiting their acceleration potential during inference. To address this, we propose a hardware-algorithm co-design framework incorporating two innovations: (1) a CIM-Aware Architecture, CIM-NET, optimized for large receptive field operation and CIM’s crossbar-based MVM acceleration; and (2) a pseudo-convolutional operator, CIM-CONV, used within CIM-NET to integrate slide-based processing with fully connected transformations for high-quality feature extraction and reconstruction. This framework reduces the number of MVM operations, improving inference speed on CIM chips while maintaining competitive performance. Experimental results indicate that, compared to the conventional lightweight model FastDVDnet, our improved variant, CIM-NET, built upon it substantially reduces MVM operations with a slight decrease in denoising performance. With a stride value of 8, CIM-NET reduces MVM operations to 1/77th of the original, while maintaining competitive PSNR (35.03 dB vs. 35.57 dB of FastDVDnet).
Förderung
- China Mobile Communications Group Company
Zitieren
APA
Gao, S., Wu, Z., Niu, Y., Li, X., & Xu, Q. (2026). CIM-NET: A Video Denoising Deep Neural Network Model Optimized for Computing-in-Memory Architectures. IEEE Signal Processing Letters, 33, 1591-1595. https://doi.org/10.1109/lsp.2026.3680452
Vancouver
Gao S, Wu Z, Niu Y, Li X, Xu Q. CIM-NET: A Video Denoising Deep Neural Network Model Optimized for Computing-in-Memory Architectures. IEEE Signal Processing Letters. 2026;33:1591-1595. doi:10.1109/lsp.2026.3680452
BibTeX
@article{Gao2026,
author = {Gao, Shan and Wu, Zhiqiang and Niu, Yawen and Li, Xiaotao and Xu, Qingqing},
title = {CIM-NET: A Video Denoising Deep Neural Network Model Optimized for Computing-in-Memory Architectures},
journal = {IEEE Signal Processing Letters},
year = {2026},
volume = {33},
pages = {1591--1595},
doi = {10.1109/lsp.2026.3680452}
}
EUVIMED