Caltech Bird Detector

Course Q&A TensorFlow: Advanced Techniques Specialization Advanced Computer Vision with TensorFlow
1

I have multiple doubts:

  1. What is the use of the 3rd bounding_boxes_on_image_array function or the 2nd function?
  2. What is the number n in the dimension s of boxes [N,4]
  3. In read_image_tfds why do we divide by 127.5 and then subtract -1 from image?
  4. Why is training_dataset an object but visualization_training_dataset used for length measurements?
  5. The boxes dimension does not signify if its ymin then xmin and so on
2

Please find my reply below.

  1. What is the use of the 3rd bounding_boxes_on_image_array function or the 2nd function? - utility function to draw a bounding box (a 2d numpy array) on image
  2. What is the number n in the dimension s of boxes [N,4] - N is the number of images
  3. In read_image_tfds why do we divide by 127.5 and then subtract -1 from image? - This is to normalize the original image to the desired scale.
  4. Why is training_dataset an object but visualization_training_dataset used for length measurements? - Could you clarity the question? I didn’t find the function visualization_training_dataset.
  5. The boxes dimension does not signify if its ymin then xmin and so on - Could you rephrase the question?
1 Like