https://github.com/edisonslightbulbs/ptod-model-training
Transfer learning using PyTorch and YOLOv5
https://github.com/edisonslightbulbs/ptod-model-training
cxx11 deep-learning k4a object-detection opencv python pytorch transfer-learning yolov5
Last synced: 3 months ago
JSON representation
Transfer learning using PyTorch and YOLOv5
- Host: GitHub
- URL: https://github.com/edisonslightbulbs/ptod-model-training
- Owner: edisonslightbulbs
- Created: 2021-09-02T12:24:14.000Z (over 3 years ago)
- Default Branch: main
- Last Pushed: 2022-03-19T18:02:05.000Z (about 3 years ago)
- Last Synced: 2025-01-14T09:12:14.943Z (4 months ago)
- Topics: cxx11, deep-learning, k4a, object-detection, opencv, python, pytorch, transfer-learning, yolov5
- Language: Jupyter Notebook
- Homepage:
- Size: 15.4 MB
- Stars: 1
- Watchers: 0
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Transfer learning using PyTorch's object detection API
| Platform | Hardware | Dependencies |
|--- |--- |--- |
| :white_square_button: Linux | :white_square_button: Azure Kinect | :white_square_button: [ gflags](https://github.com/gflags/gflags) |
|| | :white_square_button: [ glog ](https://github.com/google/glog) |
||| :white_square_button: [ Azure Kinect SDK ](https://github.com/microsoft/Azure-Kinect-Sensor-SDK) |
||| :white_square_button: [ opencv ](https://github.com/opencv/opencv) |
||| :white_square_button: [ Anaconda ](https://www.anaconda.com/products/individual) |
||| :white_square_button: [ Yolov5 ](https://github.com/ultralytics/yolov5) |
||| :white_square_button: [ Image annotation tool ](https://github.com/tzutalin/labelImg) |
---
This project is made up of two sub-projects: [`image-capturing`](./image-capturing) [`model-training`](./model-training). [`image-capturing`](./image-capturing) is a CMake project that uses Microsoft's Azure Kinect to capture so-called depth color images (of cause, this can be changed). [`model-training`](./model-training) uses shell and python scripts to exploit Tensor Flow's object detection API and train an object detection.
---
The notebooks in [`model-training README.md`](./model-training/README.md) are self-documenting, but more on that in the [`model-training README.md`](./model-training/README.md). In principle, one can use any other camera, a webcam, or even already captured images (i.e., given a reasonable number of image captures exist) to train the detection model.