Summary of Large Selective Kernel Network For Remote Sensing Object Detection, by Yuxuan Li et al.
Large Selective Kernel Network for Remote Sensing Object Detection
by Yuxuan Li, Qibin Hou, Zhaohui Zheng, Ming-Ming Cheng, Jian Yang, Xiang Li
First submitted to arxiv on: 16 Mar 2023
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: None
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 Large Selective Kernel Network (LSKNet) improves remote sensing object detection by incorporating unique prior knowledge from the scenario. Traditional methods focus on oriented bounding boxes, but LSKNet dynamically adjusts its spatial receptive field to model ranging contexts for various objects. This novel approach sets new state-of-the-art scores on HRSC2016, DOTA-v1.0, and FAIR1M-v1.0 benchmarks, achieving 98.46%, 81.85%, and 47.87% mAP respectively. The LSKNet also ranked 2nd place in the Greater Bay Area International Algorithm Competition. This paper’s contribution lies in exploring large and selective kernel mechanisms for remote sensing object detection. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The researchers created a new way to detect objects in pictures taken from far away, like satellites. They wanted to make sure their method worked well even when there are lots of small things that can be easily mistaken for bigger things. To do this, they designed a special network called LSKNet that can adjust its view to fit the size and distance of different objects. This helps it detect things more accurately. The new method is better than others at detecting objects in pictures taken from far away, and it even did well in a competition with other teams. |
Keywords
* Artificial intelligence * Object detection
