Dogs Emotion Recognition and Parameter Analysis Based on EfficientNet with Transfer Learning

Yinglin Xie

2023

Abstract

Emotion recognition in dogs plays a crucial role in understanding their mental state and improving their welfare. This study aims to develop an efficient model for recognizing emotions in dogs’ images using the Efficient Neural Network(EfficientNet) architecture. Specifically, this study uses a dataset of dog images labeled with different emotion categories. The dataset is preprocessed, while transfer learning techniques are used for training model. Fine-tuning of hyperparameters is performed to optimize the model. Finally, the trained model is evaluated on the testing set. This study is conducted on the dog-emotion-prediction dataset, achieving an accuracy of 78%. Experimental results demonstrate the effectiveness of using EfficientNet for this task. Recognizing emotions in dogs’ images has practical implications in various domains, such as veterinary care, animal behavior analysis, and pet training. It can aid veterinarians in diagnosing and treating emotional distress in dogs, assist trainers in developing tailored behavior modification strategies, and promote a deeper understanding of dogs’ emotions to strengthen the human-animal bond.

Paper Citation

in Harvard Style

Xie Y. (2023). Dogs Emotion Recognition and Parameter Analysis Based on EfficientNet with Transfer Learning. In Proceedings of the 1st International Conference on Data Analysis and Machine Learning - Volume 1: DAML; ISBN 978-989-758-705-4, SciTePress, pages 240-245. DOI: 10.5220/0012799600003885

in Bibtex Style

@conference{daml23,
author={Yinglin Xie},
title={Dogs Emotion Recognition and Parameter Analysis Based on EfficientNet with Transfer Learning},
booktitle={Proceedings of the 1st International Conference on Data Analysis and Machine Learning - Volume 1: DAML},
year={2023},
pages={240-245},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012799600003885},
isbn={978-989-758-705-4},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Data Analysis and Machine Learning - Volume 1: DAML
TI - Dogs Emotion Recognition and Parameter Analysis Based on EfficientNet with Transfer Learning
SN - 978-989-758-705-4
AU - Xie Y.
PY - 2023
SP - 240
EP - 245
DO - 10.5220/0012799600003885
PB - SciTePress