did anyone know how to train a Faster R-CNN detector to detect multiple object classes?

i wanna to train a Faster R-CNN detector to detect multiple object classes. but from document just some example train a Faster R-CNN detector to detect sigle object classes. hava no idea how to after try something hard.

Answer 1

[Harshita Gupta]

Multiple objects can be detected by formatting the input "trainingData" to contain information about multiple object classes. If you refer to the documentation:

https://www.mathworks.com/help/vision/ref/trainfasterrcnnobjectdetector.html#bvkk009-1-trainingData

you will note that the labeled ground truth "trainingData" has to be formatted as:

Labeled ground truth images, specified as a table with two or more columns. The first column must contain paths and file names to grayscale or truecolor (RGB) images. The remaining columns must contain bounding boxes related to the corresponding image. Each column represents a single object class, such as a car, dog, flower, or stop sign."

In the example linked above the "trainingData" consists of only two columns with the second column representing the object class vehicle. For detecting multiple objects, you have to label your images for the new objects you want to detect and modify the "trainingData" accordingly.

[dhruv khatri]

What if we have multiple bounding boxes in the same ground truth image? Like for a single image, multplie stop signs. How should be format our training data then?

[Pawel Habrzyk]

Bump