Finally, the Remote Sensing course discusses general engineering principles as well as history of imaging and remote sensing over time. These supplemental topics build the students’ foundations in engineering as they apply their knowledge of crisis response to class projects.
Topics Covered
- What are some classical engineering failures? What went wrong?
- What is uncertainty? How can we model and characterize uncertainty?
- What is classification? What are some techniques for classification (decision trees, K-NN)?
- What is regression? What are some techniques for regression (linear, kernel)?
- What is uncertainity? How do we characterize it (bootstrap, random forest, Gaussian models)?
- History of image recognition
- Imagenet large scale visual recognition challenge
- Convolutional neural networks
- image labeling tasks
- existing datasets
- Labeling images
- Types of annotation: scene, point, polygon
- Need for labels/data set for HADR
- Manually labeling data
- Pretrained image labelers
- Open source implementations
- Commercial implementations
- Transfer learning
- Using labels in decision making