Switch from TFA resampler op to TFA interpolate_bilinear. #669
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Switch from TFA resampler op to TFA interpolate_bilinear.
Maintaining the custom resampler op is a maintenance headache that requires aligning the versions of TFA and TFG for the opensource releases (e.g., #650). The interpolate_bilinear function is a pure-TF op that is likely somewhat slower, but has no maintenance problems. Using pure-TF ops also allows using higher-order derivatives and tf.ForwardAccumulator with the transform.sample function.
The implemented border behavior of interpolate_bilinear is DUPLICATE, whereas resampler is ZERO, so the sample function changes slightly to pad the image with zeros when BorderType.ZERO is requested.