TY - GEN
T1 - Improving Tiny YOLO with Fewer Model Parameters
AU - Liu, Yanwei
AU - Ma, Ching Wen
N1 - Publisher Copyright:
© 2021 IEEE.
PY - 2021
Y1 - 2021
N2 - With the rapid development of convolutional neural networks (CNNs), there are a variety of techniques that can improve existing CNN models, including attention mechanisms, activation functions, and data augmentation. However, integrating these techniques can lead to a significant increase in the number of parameters and FLOPs. Here, we integrated Efficient Channel Attention Net(ECA-Net), Mish activation function, All Convolutional Net (ALL-CNN), and a twin detection head architecture into YOLOv4-tiny, yielding an AP50 of 44.2% on the MS COCO 2017 dataset. The proposed Attention ALL-CNN Twin Head YOLO (A2-YOLO) outperforms the original YOLOv4-tiny on the same dataset by 3.3% and reduces the model parameters by 7.26%. Source code is at https://github.com/e96031413/AA-YOLO
AB - With the rapid development of convolutional neural networks (CNNs), there are a variety of techniques that can improve existing CNN models, including attention mechanisms, activation functions, and data augmentation. However, integrating these techniques can lead to a significant increase in the number of parameters and FLOPs. Here, we integrated Efficient Channel Attention Net(ECA-Net), Mish activation function, All Convolutional Net (ALL-CNN), and a twin detection head architecture into YOLOv4-tiny, yielding an AP50 of 44.2% on the MS COCO 2017 dataset. The proposed Attention ALL-CNN Twin Head YOLO (A2-YOLO) outperforms the original YOLOv4-tiny on the same dataset by 3.3% and reduces the model parameters by 7.26%. Source code is at https://github.com/e96031413/AA-YOLO
KW - Deep Learning
KW - Object Detection
KW - YOLO
UR - http://www.scopus.com/inward/record.url?scp=85124035419&partnerID=8YFLogxK
U2 - 10.1109/BigMM52142.2021.00017
DO - 10.1109/BigMM52142.2021.00017
M3 - Conference contribution
AN - SCOPUS:85124035419
T3 - Proceedings - 2021 IEEE 7th International Conference on Multimedia Big Data, BigMM 2021
SP - 61
EP - 64
BT - Proceedings - 2021 IEEE 7th International Conference on Multimedia Big Data, BigMM 2021
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 7th IEEE International Conference on Multimedia Big Data, BigMM 2021
Y2 - 15 November 2021 through 17 November 2021
ER -