bounding_box.to_ragged() fails with keras 3.

[smitlg]

Steps to reproduce:

>>> bounding_boxes
{'classes': <tf.RaggedTensor [[5.0], [5.0, 5.0], [4.0, 5.0], [5.0, 5.0, 5.0, 5.0]]>,
 'boxes': <tf.RaggedTensor [[[331.08414, 0.0, 537.0, 56.885612]],
  [[283.4905, 132.49506, 349.13153, 318.0734],
   [217.2055, 164.95471, 295.0306, 338.5343]] ,
  [[305.12332, 426.07184, 640.0, 640.0],
   [307.50845, 517.3163, 334.21893, 582.8346]],
  [[244.3948, 0.0, 423.6596, 349.30737],
   [433.71707, 0.0, 479.48785, 26.02331],
   [310.58087, 0.0, 347.29642, 48.72561],
   [164.48027, 0.0, 282.78375, 236.52121]]]>}
>>> b_dense = bounding_box.to_dense(bounding_boxes)
>>> b_dense
{'classes': <tf.Tensor: shape=(4, 4), dtype=float32, numpy=
 array([[ 5., -1., -1., -1.],
        [ 5.,  5., -1., -1.],
        [ 4.,  5., -1., -1.],
        [ 5.,  5.,  5.,  5.]], dtype=float32)>,
 'boxes': <tf.Tensor: shape=(4, 4, 4), dtype=float32, numpy=
 array([[[331.08414 ,   0.      , 537.      ,  56.885612],
         [ -1.      ,  -1.      ,  -1.      ,  -1.      ],
         [ -1.      ,  -1.      ,  -1.      ,  -1.      ],
         [ -1.      ,  -1.      ,  -1.      ,  -1.      ]],
 
        [[283.4905  , 132.49506 , 349.13153 , 318.0734  ],
         [217.2055  , 164.95471 , 295.0306  , 338.5343  ],
         [ -1.      ,  -1.      ,  -1.      ,  -1.      ],
         [ -1.      ,  -1.      ,  -1.      ,  -1.      ]],
 
        [[305.12332 , 426.07184 , 640.      , 640.      ],
         [307.50845 , 517.3163  , 334.21893 , 582.8346  ],
         [ -1.      ,  -1.      ,  -1.      ,  -1.      ],
         [ -1.      ,  -1.      ,  -1.      ,  -1.      ]],
 
        [[244.3948  ,   0.      , 423.6596  , 349.30737 ],
         [433.71707 ,   0.      , 479.48785 ,  26.02331 ],
         [310.58087 ,   0.      , 347.29642 ,  48.72561 ],
         [164.48027 ,   0.      , 282.78375 , 236.52121 ]]], dtype=float32)>}
>>> b_ragged = bounding_box.to_ragged(b_dense)
NotImplementedError                       Traceback (most recent call last)
[<ipython-input-23-a99e42282bd8>](https://localhost:8080/#) in <cell line: 1>()
----> 1 b_ragged = bounding_box.to_ragged(b_dense)
      2 b_ragged

[/usr/local/lib/python3.10/dist-packages/keras_cv/bounding_box/to_ragged.py](https://localhost:8080/#) in to_ragged(bounding_boxes, sentinel, dtype)
     55     """
     56     if backend.supports_ragged() is False:
---> 57         raise NotImplementedError(
     58             "`bounding_box.to_ragged` was called using a backend which does "
     59             "not support ragged tensors. "

NotImplementedError: `bounding_box.to_ragged` was called using a backend which does not support ragged tensors. Current backend: tensorflow.

Keras Info:

>>> keras.version(), keras.backend.backend()
('3.1.1', 'tensorflow')

Keras CV is installed using the current master branch.

[sachinprasadhs]

RaggedTensor is not yet implemented yet, in future Keras team is planning to support it only for TensorFlow backend, till then you can refer the suggestion here to use dynamic input keras-team/keras#18414 (comment)

[github-actions]

This issue is stale because it has been open for 14 days with no activity. It will be closed if no further activity occurs. Thank you.

[github-actions]

This issue was closed because it has been inactive for 28 days. Please reopen if you'd like to work on this further.