Custom hardware architectures for deep learning on portable devices: a review

KS Zaman, MBI Reaz, SHM Ali… - … on Neural Networks …, 2021 - ieeexplore.ieee.org
The staggering innovations and emergence of numerous deep learning (DL) applications
have forced researchers to reconsider hardware architecture to accommodate fast and …

Design and analysis of approximate compressors for balanced error accumulation in MAC operator

G Park, J Kung, Y Lee - … Transactions on Circuits and Systems I …, 2021 - ieeexplore.ieee.org
In this paper, we present a novel approximate computing scheme suitable for realizing the
energy-efficient multiply-accumulate (MAC) processing. In contrast to the prior works that …

CAP-YOLO: Channel attention based pruning YOLO for coal mine real-time intelligent monitoring

Z Xu, J Li, Y Meng, X Zhang - Sensors, 2022 - mdpi.com
Real-time coal mine intelligent monitoring for pedestrian identifying and positioning is an
important means to ensure safety in production. Traditional object detection models based …

Approximate LSTM computing for energy-efficient speech recognition

J Jo, J Kung, Y Lee - Electronics, 2020 - mdpi.com
This paper presents an approximate computing method of long short-term memory (LSTM)
operations for energy-efficient end-to-end speech recognition. We newly introduce the …

Diagonal-kernel convolutional neural networks for image classification

G Li, X Shen, J Li, J Wang - Digital Signal Processing, 2021 - Elsevier
The recognition performance of convolutional neural networks has surpassed that of
humans in many computer vision areas. However, there is a large number of parameter …

Transforming agriculture with Machine Learning, Deep Learning, and IoT: perspectives from Ethiopia—challenges and opportunities

NE Benti, MD Chaka, AG Semie, B Warkineh… - Discover Agriculture, 2024 - Springer
Agriculture holds a crucial position in maintaining livelihoods and securing food sources,
particularly in nations such as Ethiopia, where a substantial portion of the population …

Effects of hidden layer sizing on CNN fine-tuning

S Marrone, C Papa, C Sansone - Future Generation Computer Systems, 2021 - Elsevier
Some applications have the property of being resilient, meaning that they are robust to noise
(eg due to error) in the data. This characteristic is very useful in situations where an …

Lightweight end-to-end stress recognition using binarized CNN-LSTM models

M Yun, S Hong, S Yoo, J Kim… - 2022 IEEE 4th …, 2022 - ieeexplore.ieee.org
In this paper, we propose a novel end-to-end stress recognition model by combining
binarized convolutional neural network (CNN) and long short-term memory (LSTM) models …

Performance evaluation of compressed deep CNN for motor imagery classification using EEG

R Vishnupriya, N Robinson, R Reddy… - 2021 43rd Annual …, 2021 - ieeexplore.ieee.org
Recently, deep learning and convolutional neural networks (CNNs) have reported several
promising results in the classification of Motor Imagery (MI) using Electroencephalography …

Energy-efficient risc-v-based vector processor for cache-aware structurally-pruned transformers

JG Min, D Kam, Y Byun, G Park… - 2023 IEEE/ACM …, 2023 - ieeexplore.ieee.org
Based on recent RISC-V designs, we present in this paper a low-power vector processor
architecture for efficiently deploying vision transformer (ViT) models. To fairly measure the …