Thannks to https://github.com/Tony607/object_detection_demo for the intial demo!
- A tutorial to train and use MobileNetSSDv2 with the TensorFlow Object Detection API
- A tutorial to train and use Faster R-CNN with the TensorFlow Object Detection API
- How to load your custom image detection from Roboflow (here we use a public blood cell dataset with tfrecord)
- Download base MobileNetSSDv2 model
- Set up training environment
- Configure training pipeline and train the model
- Export the trained model's .pb inference graph
- Use the saved model for inference
- How to load your custom image data from Roboflow (here we use a public blood cell dataset with tfrecord)
- Download base pretrained Faster R-CNN model
- Set up training environment
- Configure training pipeline and train model
- Export the trained model's .pb inference graph
- Use the saved model for inference
- This blog post for MobileNetSSDv2 walks through the tutorial
- This blog post for Faster R-CNN walks through the tutorial
- For the MobileNetSSDv2 model tutorial
- For the Faster R-CNN model tutorial
- For reading purposes, for MobileNetSSDv2, the notebook is saved here as Tutorial_Mobilenet.ipynb
- For reading purposes, for Faster R-CNN, the notebook is also saved here as Tutorial_Faster_RCNN.ipynb
Roboflow makes managing, preprocessing, augmenting, and versioning datasets for computer vision seamless. Developers reduce 50% of their code when using Roboflow's workflow, automate annotation quality assurance, save training time, and increase model reproducibility.
