I have kitti dataset for training my Yolo model. The training dataset of kiiti dataset contains images with multiple objects. How to train that ?
Intuitively, I thought to use single object images like to train the model for car and pedestrian one by one with different images, but in Kitti dataset, all the objects are in single images
Anyone who has worked with Kitti dataset ?
Have a look at the Detection tutorial, where the target is explained for this kind of task.
In the Defining a Dataset section you can see that multiple boxes are provided for each image. While this tutorial uses another dataset, you might need to adapt this work flow to your dataset.
Thanks. I got it. I was little confused over the terminologies of anchor boxes and bounding boxes given in online blogs. So, I was not sure to train multiple objects in an image.