Unckecked rank size in `tf.image.generate_bounding_box_proposals` in Tensorflow
tensorflow
Affected versions
| Product | Affected | Fixed in |
|---|---|---|
| tensorflow PyPI |
|
|
| tensorflow-cpu PyPI |
|
tensorflow
| Product | Affected | Fixed in |
|---|---|---|
| tensorflow PyPI |
|
|
| tensorflow-cpu PyPI |
|
≥ 2.9.0 · < 2.9.3≥ 2.10.0 · < 2.10.1
|
| tensorflow-gpu PyPI |
|
|
TensorFlow is an open source platform for machine learning. When running on GPU, `tf.image.generate_bounding_box_proposals` receives a `scores` input that must be of rank 4 but is not checked. We have patched the issue in GitHub commit cf35502463a88ca7185a99daa7031df60b3c1c98. The fix will be included in TensorFlow 2.11. We will also cherrypick this commit on TensorFlow 2.10.1, 2.9.3, and TensorFlow 2.8.4, as these are also affected and still in supported range.
From the CVE record · GitHub_M
When running on a GPU, TensorFlow’s `tf.image.generate_bounding_box_proposals` does not check that its `scores` input has the required rank of 4. This allows an incorrectly ranked input to reach the operation; the input does not specify a resulting exploit or whether an attacker needs network access, authentication, or user interaction.
Engineers running PyPI `tensorflow`, `tensorflow-cpu`, or `tensorflow-gpu` versions below 2.8.4, from 2.9.0 to below 2.9.3, or from 2.10.0 to below 2.10.1 should check their installations.