Few-Shot Learning with Multi-Scale Feature Fusion for Microalgae Classification

Guihong Yuan

doi:10.6919/ICJE.202502_11(2).0021

Authors

Guihong Yuan

DOI:

https://doi.org/10.6919/ICJE.202502_11(2).0021

Keywords:

Phytoplankton Microalgae; Multi-Scale Feature Fusion; Few-Shot Learning; Metric Learning.

Abstract

Phytoplankton plays a multifaceted role in the field of aquaculture, serving not only as an important food source for farmed organisms but also rich in bioactive compounds and nutrients. Due to their sensitivity to changes in the aquatic environment, phytoplankton are crucial biological indicators for assessing water quality. Traditionally, the identification of phytoplankton relies on the manual collection of samples and professional analysis under a microscope, a process that is both time-consuming and labor-intensive, and requires a high level of professional expertise and identification skills from the inspectors. Additionally, the recognition of the subtle features of microalgae is quite challenging. To address this issue, this study proposes a few-shot learning algorithm that integrates multi-scale features, aiming to enhance the efficiency of feature extraction for microalgae. Furthermore, the algorithm incorporates metric learning techniques. Under the experimental setup of 5-way 5-shot, the classification accuracy of this method reached 80.51%, which is a 3.06% improvement over the previous suboptimal model.

Downloads

Download data is not yet available.

References

[1] Mugani R, El Khalloufi F, Kasada M, et al. Monitoring of toxic cyanobacterial blooms in Lalla Takerkoust reservoir by satellite imagery and microcystin transfer to surrounding farms[J]. Harmful Algae, 2024, 135: 102631.

[2] Chorus I, Fastner J, Welker M. Cyanobacteria and cyanotoxins in a changing environment: Concepts, controversies, challenges[J]. Water, 2021, 13(18): 2463.

[3] Liu J Y, Zeng L H, Ren Z H. Recent application of spectroscopy for the detection of microalgae life information: A review[J]. Applied Spectroscopy Reviews, 2020, 55(1): 26-59.

[4] Li X Y, Yu R C, Richardson A J, et al. Marked shifts of harmful algal blooms in the Bohai Sea linked with combined impacts of environmental changes[J]. Harmful Algae, 2023, 121: 102370.

[5] Song Y, Wang T, Cai P, et al. A comprehensive survey of few-shot learning: Evolution, applications, challenges, and opportunities[J]. ACM Computing Surveys, 2023, 55(13s): 1-40.

[6] Snell J, Swersky K, Zemel R. Prototypical networks for few-shot learning[J]. Advances in neural information processing systems, 2017, 30.

[7] Sung F, Yang Y, Zhang L, et al. Learning to compare: Relation network for few-shot learning[C]// Proceedings of the IEEE conference on computer vision and pattern recognition. 2018: 1199-1208.

[8] Tian Y, Wang Y, Krishnan D, et al. Rethinking few-shot image classification: a good embedding is all you need?[C]//Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XIV 16. Springer International Publishing, 2020: 266-282.

[9] Zhang C, Cai Y, Lin G, et al. Deepemd: Few-shot image classification with differentiable earth mover's distance and structured classifiers[C]//Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2020: 12203-12213.

[10] Zhang H, Zu K, Lu J, et al. Epsanet: An efficient pyramid split attention block on convolutional neural network[J]. CoRR, 2021.

[11] Xie J, Long F, Lv J, et al. Joint distribution matters: Deep brownian distance covariance for few-shot classification[C]//Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2022: 7972-7981.

[12] Wah, C., Branson, S., Welinder, P., Perona, P., and Belongie, S.J. 2011. The Caltech-UCSD Birds-200-2011 Dataset.

[13] Chen C F R, Fan Q, Panda R. Crossvit: Cross-attention multi-scale vision transformer for image classification[C]//Proceedings of the IEEE/CVF international conference on computer vision. 2021: 357-366.

[14] Székely G J, Rizzo M L. Brownian distance covariance[J]. 2009.

[15] Székely G J, Rizzo M L, Bakirov N K. Measuring and testing dependence by correlation of distances[J]. 2007.

[16] Zheng Y, Liu S, Chen H, et al. Hybrid FusionNet: A hybrid feature fusion framework for multi-source high-resolution remote sensing image classification[J]. IEEE Transactions on Geoscience and Remote Sensing, 2024.