ACLNet: An Attention and Clustering-based Cloud Segmentation Network
We propose a novel deep learning model named ACLNet, for cloud segmentation from ground images. ACLNet uses both deep neural network and machine learning (ML) algorithm to extract complementary features. Specifically, it uses EfficientNet-B0 as the backbone, "`a trous spatial pyramid pooling" (ASPP) to learn at multiple receptive fields, and "global attention module" (GAM) to extract finegrained details from the image. ACLNet also uses k-means clustering to extract cloud boundaries more precisely. ACLNet is effective for both daytime and nighttime images. It provides lower error rate, higher recall and higher F1-score than state-of-art cloud segmentation models. The source-code of ACLNet is available here: https://github.com/ckmvigil/ACLNet.
PDF Abstract
SWIMSEG
SWINSEG
SWINySEG
| Task | Dataset | Model | Metric Name | Metric Value | Global Rank | Benchmark |
|---|---|---|---|---|---|---|
| Semantic Segmentation | SWIMSEG | ACLNet | Average Precision | 0.964 | # 1 | |
| Average Recall | 0.979 | # 1 | ||||
| F1-Score | 0.971 | # 1 | ||||
| Mean IoU | 0.992 | # 1 | ||||
| MCC | 0.956 | # 1 | ||||
| Semantic Segmentation | SWINSEG | ACLNet | Average Precision | 0.917 | # 1 | |
| Average Recall | 0.982 | # 1 | ||||
| F1-Score | 0.947 | # 1 | ||||
| Mean IoU | 0.985 | # 1 | ||||
| MCC | 0.930 | # 1 | ||||
| Semantic Segmentation | SWINySEG | ACLNet | Average Precision | 0.959 | # 1 | |
| Average Recall | 0.979 | # 1 | ||||
| F1-Score | 0.968 | # 1 | ||||
| Mean IoU | 0.993 | # 1 | ||||
| MCC | 0.960 | # 1 |
