Pinned Repositories
D-LinkNet
DeepGlobe-Road-Extraction-Challenge
PaRK-Detect
PaRK-Detect: Towards Efficient Multi-Task Satellite Imagery Road Extraction via Patch-Wise Keypoints Detection
Stagewise-Weakly-Supervised-Satellite-Imagery-Road-Extraction-Based-on-Road-Centerline-Scribbles
Satellite imagery semantic segmentation requires high-resolution satellite images to train the Nerual Network. Due to the expensive cost of the pixel-level annotation, the cheaper instance-level annotation of the satellite image is not suitable for the high-resolution image, which leads to the high error rate. Driven by the need to avoid a large number of false detections and missed detections with coarse labels, the weakly-supervised learning holds the promise of data exploration with for example the learning method capable of refining the mask with the location of the seed. In addition, this project needs to generate the more specific label's information as the output of the designed weakly-supervised algorithm. This project is looking for someone who has a basic knowledge of deep learning and a good skill for programming with python.
ShenweiXie's Repositories
ShenweiXie/PaRK-Detect
PaRK-Detect: Towards Efficient Multi-Task Satellite Imagery Road Extraction via Patch-Wise Keypoints Detection
ShenweiXie/D-LinkNet
DeepGlobe-Road-Extraction-Challenge
ShenweiXie/Stagewise-Weakly-Supervised-Satellite-Imagery-Road-Extraction-Based-on-Road-Centerline-Scribbles
Satellite imagery semantic segmentation requires high-resolution satellite images to train the Nerual Network. Due to the expensive cost of the pixel-level annotation, the cheaper instance-level annotation of the satellite image is not suitable for the high-resolution image, which leads to the high error rate. Driven by the need to avoid a large number of false detections and missed detections with coarse labels, the weakly-supervised learning holds the promise of data exploration with for example the learning method capable of refining the mask with the location of the seed. In addition, this project needs to generate the more specific label's information as the output of the designed weakly-supervised algorithm. This project is looking for someone who has a basic knowledge of deep learning and a good skill for programming with python.