Weakly Supervised Collaborative Learning for Airborne Pollen Segmentation and Classification from SEM Images
Open Access
Abstract
Existing pollen identification methods heavily rely on the scale and quality of pollen images. However, there are many impurities in real-world SEM images that should be considered. This paper proposes a collaborative learning method to jointly improve the performance of pollen segmentation and classification in a weakly supervised manner. It first locates pollen regions from the raw images based on the detection model. To improve the classification performance, we segmented the pollen grains through a pre-trained U-Net using unsupervised pollen contour features. The segmented pollen regions were fed into a deep convolutional neural network to obtain the activation maps, which were used to further refine the segmentation masks. In this way, both segmentation and classification models can be collaboratively trained, supervised by just pollen contour features and class-specific information. Extensive experiments on real-world datasets were conducted, and the results prove that our method effectively avoids impurity interference and improves pollen identification accuracy (86.6%) under the limited supervision (around 1000 images with image-level labels).Keywords
Funding Information
- Beijing Municipal Science and Technology Project (Z191100009119013)
This publication has 36 references indexed in Scilit:
- Analysis of the pollen allergen content of twelve olive cultivars grown in PortugalAerobiologia, 2013
- Aerobiology and Its Role in the Transmission of Infectious DiseasesJournal of Pathogens, 2013
- Classifying black and white spruce pollen using layered machine learningNew Phytologist, 2012
- Feasibility study on automated recognition of allergenic pollen: grass, birch and mugwortAerobiologia, 2006
- Towards automation of palynology 2: the use of texture measures and neural network analysis for automated identification of optical images of pollen grainsJournal of Quaternary Science, 2004
- Towards automation of palynology 1: analysis of pollen shape and ornamentation using simple geometric measures, derived from scanning electron microscope imagesJournal of Quaternary Science, 2004
- Towards automation of palynology 3: pollen pattern recognition using Gabor transforms and digital momentsJournal of Quaternary Science, 2004
- COLOUR IMAGE IN 2D AND 3D MICROSCOPY FOR THE AUTOMATION OF POLLEN RATE MEASUREMENTImage Analysis and Stereology, 2002
- A new approach to automated pollen analysisQuaternary Science Reviews, 2000
- Computerized identification of pollen grains by texture analysisReview of Palaeobotany and Palynology, 1990