Summary of Learning Without Exact Guidance: Updating Large-scale High-resolution Land Cover Maps From Low-resolution Historical Labels, by Zhuohong Li et al.
Learning without Exact Guidance: Updating Large-scale High-resolution Land Cover Maps from Low-resolution Historical Labels
by Zhuohong Li, Wei He, Jiepan Li, Fangxiao Lu, Hongyan Zhang
First submitted to arxiv on: 5 Mar 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed Paraformer framework leverages a parallel CNN-Transformer feature extractor to efficiently guide large-scale high-resolution (HR) land-cover mapping using easy-access historical low-resolution (LR) data. The approach combines the strengths of convolutional neural networks (CNNs) in preserving local details with transformer-based global modeling, effectively capturing both contextual information. To address spatial mismatch between training data and HR images, a pseudo-label-assisted training (PLAT) module is employed to refine LR labels for weakly supervised semantic segmentation. Experimental results on two large-scale datasets demonstrate the superiority of Paraformer over state-of-the-art methods in updating HR land-cover maps from LR historical labels. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Paraformer is a new way to create high-resolution maps of the Earth’s surface using old, low-resolution data. This helps solve big challenges like climate change and natural disasters. The problem is that making these maps is hard because it needs to show lots of details and cover large areas. Paraformer uses a special combination of two types of computer vision models: CNNs and transformers. These models work together to capture both small details and big patterns. To make sure the results are accurate, Paraformer also refines old labels using new information. The results show that Paraformer is better than other methods at creating these maps. |
Keywords
* Artificial intelligence * Cnn * Semantic segmentation * Supervised * Transformer
