TY - GEN
T1 - A Classification Method for Small Sample Multi-label Images
AU - Li, Ruohan
AU - Jiang, Zengru
AU - Dai, Wei
AU - Nie, Yongkang
AU - Liu, Liang
AU - Dai, Yaping
N1 - Publisher Copyright:
© 2019 IEEE.
PY - 2019/6
Y1 - 2019/6
N2 - This paper studies the classification problem of the small sample multi-label image scene recognition. Combining convolutional neural network (CNN) and multi-label K neighborhood algorithm (MLKNN), the CNN-MLKNN classification method is proposed. The method uses CNN to automatically extract the features of small sample images, and combines transfer learning to optimize the model structure and weight to reduce the risk of over-fitting. MLKNN algorithm is used to replace the sigmoid function of CNN, and the output features of the FC layer are used as input features of MLKNN for image classifier training. Based on the classification experiments of two small sample multi-label image sets, seven multi-label evaluation indicators are used for testing. The experimental results show that the CNN-MLKNN method proposed in this paper has a better classification effect.
AB - This paper studies the classification problem of the small sample multi-label image scene recognition. Combining convolutional neural network (CNN) and multi-label K neighborhood algorithm (MLKNN), the CNN-MLKNN classification method is proposed. The method uses CNN to automatically extract the features of small sample images, and combines transfer learning to optimize the model structure and weight to reduce the risk of over-fitting. MLKNN algorithm is used to replace the sigmoid function of CNN, and the output features of the FC layer are used as input features of MLKNN for image classifier training. Based on the classification experiments of two small sample multi-label image sets, seven multi-label evaluation indicators are used for testing. The experimental results show that the CNN-MLKNN method proposed in this paper has a better classification effect.
KW - Convolutional Neural Network
KW - Multi-label Classification
KW - Multi-label K Neighborhood Algorithm
KW - Small Sample Data
KW - Transfer Learning
UR - https://www.scopus.com/pages/publications/85073096742
U2 - 10.1109/CCDC.2019.8832422
DO - 10.1109/CCDC.2019.8832422
M3 - Conference contribution
AN - SCOPUS:85073096742
T3 - Proceedings of the 31st Chinese Control and Decision Conference, CCDC 2019
SP - 1365
EP - 1370
BT - Proceedings of the 31st Chinese Control and Decision Conference, CCDC 2019
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 31st Chinese Control and Decision Conference, CCDC 2019
Y2 - 3 June 2019 through 5 June 2019
ER -