Identifying the objects in images/video using the faster rcnn model from the tensorflow zoo
Using the Tensorflow document installing and verifying libraries.
After installation change the pipeline.config file according to the instructions in the tensorflow object detection document
After adding changes to the Pipeline.config file add it to the models folder and run the training model.
We can change the batch size and number of steps in the training process to increase the accuracy of the model
- More the number the steps it is trained more is the accuracy
- Time it takes to train the model increases with the number of steps.