Reconocimiento de animales desde imágenes utilizando aprendizaje por transferencia

Resumen

Los sistemas de reconocimiento automático basados en imágenes se han utilizado ampliamente para resolver diferentes tareas de visión por computador. En particular, la identificación de animales en granjas es un campo de investigación de interés para comunidad relacionada con visión artificial y agricultura. En este sentido, es necesario desarrollar algoritmos robustos y precisos para respaldar las tareas de detección, reconocimiento y monitoreo, en aras de apoyar la gestión de granjas en agricultura. Tradicionalmente, se han propuesto enfoques de aprendizaje profundo para resolver tareas de detección basadas en imágenes. No obstante, se requieren de bases de datos con muchas instancias para lograr un rendimiento competitivo, sin mencionar los problemas de ajuste de los hiperparámetros. En este artículo, proponemos un enfoque de aprendizaje por transferencia para el reconocimiento de animales basado en imágenes. En particular, mejoramos un modelo de red neuronal convolucional previamente entrenado para la clasificación de animales a partir de imágenes ruidosas y de baja calidad. Primero, se prueba una tarea de perro contra gato a partir de la conocida base de datos CIFAR. Además, se crea una base de datos de vaca versus no vaca para probar nuestro enfoque de aprendizaje por transferencia. Los resultados obtenidos muestran un rendimiento de clasificación competitivo utilizando diferentes tipos de arquitecturas, en comparación con las metodologías actuales.

Citas

[1] Qun Liu and S. Mukhopadhyay, “Unsupervised Learning using Pretrained CNN and Associative Memory Bank.” 2018. doi: 10.1109/IJCNN.2018.8489408
[2] Cai, Z., & Vasconcelos, N. (2019). Cascade R-CNN: high quality object detection and instance segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence. doi: 10.1109/TPAMI.2019.2956516
[3] Liu, C., Liu, P., Zhao, W., & Tang, X. (2019). Visual Tracking by Structurally Optimizing Pre-trained CNN. IEEE Transactions on Circuits and Systems for Video Technology. doi: 10.1109/TCSVT.2019.2938038
[4] Yao, G., Lei, T., & Zhong, J. (2019). A review of Convolutional-Neural-Network-based action recognition. Pattern Recognition Letters, 118, 14-22. doi: 10.1016/j.patrec.2018.05.018
[5] Su, J., Yi, D., Su, B., Mi, Z., Liu, C., Hu, X., ... & Chen, W. H. (2020). Aerial Visual Perception in Smart Farming: Field Study of Wheat Yellow Rust Monitoring. IEEE Transactions on Industrial Informatics. doi: 10.1109/TII.2020.2979237
[6] Mukherjee, A., Misra, S., Sukrutha, A., & Raghuwanshi, N. S. (2020). Distributed aerial processing for IoT-based edge UAV swarms in smart farming. Computer Networks, 167, 107038. doi: 10.1016/j.comnet.2019.107038
[7] Bullock, J., Cuesta-Lázaro, C., & Quera-Bofarull, A. (2019, March). XNet: A convolutional neural network (CNN) implementation for medical X-Ray image segmentation suitable for small datasets. In Medical Imaging 2019: Biomedical Applications in Molecular, Structural, and Functional Imaging (Vol. 10953, p. 109531Z). International Society for Optics and Photonics. doi: 10.1117/12.2512451
[8] Zhou, T., Ruan, S., & Canu, S. (2019). A review: Deep learning for medical image segmentation using multi-modality fusion. Array, 3, 100004. doi: 10.1016/j.array.2019.100004
[9] Xu, Q., Zhang, M., Gu, Z., & Pan, G. (2019). Overfitting remedy by sparsifying regularization on fully-connected layers of CNNs. Neurocomputing, 328, 69-74. doi: 10.1016/j.neucom.2018.03.080
[10] Webster, R., Rabin, J., Simon, L., & Jurie, F. (2019). Detecting overfitting of deep generative networks via latent recovery. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 11273-11282).
[11] Lu, N., Zhang, T., Niu, G., & Sugiyama, M. (2020, June). Mitigating overfitting in supervised classification from two unlabeled datasets: A consistent risk correction approach. In International Conference on Artificial Intelligence and Statistics (pp. 1115-1125).
[12] Li, Z., Kamnitsas, K., & Glocker, B. (2019, October). Overfitting of neural nets under class imbalance: Analysis and improvements for segmentation. In International Conference on Medical Image Computing and Computer-Assisted Intervention (pp. 402-410). Springer, Cham. doi: 10.1007/978-3-030-32248-9_98
[13] Bejani, M. M., & Ghatee, M. (2020). Theory of adaptive SVD regularization for deep neural networks. Neural Networks. doi: 10.1016/j.neunet.2020.04.021
[14] Ranjit, M. P., Ganapathy, G., Sridhar, K., & Arumugham, V. (2019, July). Efficient deep learning hyperparameter tuning using cloud infrastructure: intelligent distributed hyperparameter tuning with Bayesian optimization in the cloud. In 2019 IEEE 12th International Conference on Cloud Computing (CLOUD) (pp. 520-522). IEEE. doi: 10.1109/CLOUD.2019.00097
[15] Choi, D., Shallue, C. J., Nado, Z., Lee, J., Maddison, C. J., & Dahl, G. E. (2019). On empirical comparisons of optimizers for deep learning. arXiv preprint arXiv:1910.05446.
[16] Zhuang, F., Qi, Z., Duan, K., Xi, D., Zhu, Y., Zhu, H., ... & He, Q. (2020). A comprehensive survey on transfer learning. Proceedings of the IEEE. doi: 10.1109/JPROC.2020.3004555
[17] Zheng, H., Wang, R., Yang, Y., Yin, J., Li, Y., Li, Y., & Xu, M. (2019). Cross-domain fault diagnosis using knowledge transfer strategy: A review. IEEE Access, 7, 129260-129290. doi: 10.1109/ACCESS.2019.2939876
[18] Khan, S., Islam, N., Jan, Z., Din, I. U., & Rodrigues, J. J. C. (2019). A novel deep learning-based framework for the detection and classification of breast cancer using transfer learning. Pattern Recognition Letters, 125, 1-6. doi: 10.1016/j.patrec.2019.03.022
[19] Sun, Q., Liu, Y., Chua, T. S., & Schiele, B. (2019). Meta-transfer learning for few-shot learning. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 403-412).
[20] Sharma, N., Jain, V., & Mishra, A. (2018). An analysis of convolutional neural networks for image classification. Procedia computer science, 132, 377-384. 10.1016/j.procs.2018.05.198
[21] Liu, Q., & Mukhopadhyay, S. (2018, July). Unsupervised learning using pretrained CNN and associative memory bank. In 2018 International Joint Conference on Neural Networks (IJCNN) (pp. 01-08). IEEE. 10.1109/IJCNN.2018.8489408
[22] Géron, A. (2019). Hands-on machine learning with Scikit-Learn, Keras, and TensorFlow: Concepts, tools, and techniques to build intelligent systems. O'Reilly Media.
[23] Heaton, J. (2018). Ian goodfellow, yoshua bengio, and aaron courville: Deep learning. doi: 10.1007/s10710-017-9314-z
[24] Trnovszký, T., Kamencay, P., Orješek, R., Benčo, M., & Sýkora, P. (2017). Animal recognition system based on convolutional neural network. doi: 10.15598/aeee.v15i3.2202
[25] Nguyen, H., Maclagan, S. J., Nguyen, T. D., Nguyen, T., Flemons, P., Andrews, K., ... & Phung, D. (2017, October). Animal recognition and identification with deep convolutional neural networks for automated wildlife monitoring. In 2017 IEEE international conference on data science and advanced Analytics (DSAA) (pp. 40-49). IEEE. doi: 10.1109/DSAA.2017.31
Publicado
2021-09-30
Cómo citar
Collazos Huertas, D., Gómez Gómez, G., & Álvarez Meza, A. (2021). Reconocimiento de animales desde imágenes utilizando aprendizaje por transferencia. Scientia Et Technica, 26(03), 406-411. https://doi.org/10.22517/23447214.24538
Sección
Bioingeniería