项目作者： sshh12
项目描述 ：
Unmanned Aerial Vehicle Team @ UT Austin
高级语言： Python
项目地址: git://github.com/sshh12/UAV-Image-Rec.git
UAV Image Recognition
A snapshot of the code used by the Unmanned Aerial Vehicle Team at Austin (UAVA) for image recognition tasks.
Tech
target-finder-model
target-finder-model contains:
- Scripts to generate fake test imagery with labels for training
- A CNN model for filtering imagery that either may contain or does not contain a target
- A darknet implementation of YOLOv3 for detecting the location of targets as well as their characteristics (writing, shape, etc)
- Hyperparams, configs, training scripts for both models
target-finder
target-finder contains:
- Scripts which preprocess raw imagery for the models
- The inference code for detection and interpretation of the models output in an actionable format
- An easy to use
find_targets(image) API
Originial Repos
Primary Contributors
