This plugin provides tools to perform object detection using Deep Learning.
Object detection consists in detecting the location and class of several objects in the same image.
This plugin comes with four recipes and a macro.
Plugin Information
| Version | 0.1.7 |
|---|---|
| Author | Dataiku |
| Released | 2018-08-01 |
| Last updated | 2022-09-01 |
| License | Apache Software License |
| Source code | Github: CPU version, GPU version |
| Reporting issues | Github: CPU version, GPU version |
⚠️ Starting with DSS version 10 this plugin is partially superseded by the native object detection capabilities.
Recipes
Fine-tune detection model
This recipe does transfer learning and finetuning to adapt a pretrained model on a new dataset.
Detect objects
This recipe detects objects in images and produces a dataset storing all the detected objects with their class and localization.
Display bounding boxes
This recipe gives objects localization, draw on images a bounding box around the objects.
Detect objects in video
This recipe detects objects in a video and produces a copy of the video with the objects drawn on it.
If ffmpeg is installed the video will be of the mp4 format, else of the mkv format. To install ffmpeg, you can use the following command:
- On Ubuntu: sudo apt-get install ffmpeg
- On macOS: brew install ffmpeg