Batch support for deterministic image ops

[JoshVarty]

Based on #14854 and working on #8926

Previously I had implemented batch support for a number of image ops. However performance concerns were raised and the changes were reverted.

I have re-implemented the changes for:

• flip_left_right
• flip_up_down
• transpose_image
• rot90

I ran performance tests from #15348 with:

bazel run -c opt //tensorflow/python:image_ops_test -- --benchmarks=FlipImageBenchmark

Operation	Before (μs)	After (μs)
benchmarkFlipLeftRight_299_299_3_/cpu:0_1	274.49	264.26
benchmarkFlipLeftRight_299_299_3_/cpu:0__all	292.76	266.10
benchmarkFlipLeftRight_299_299_3___all	273.80	265.71
*benchmarkRandomFlipLeftRight_299_299_3_/cpu:0_1	242.58	241.89
*benchmarkRandomFlipLeftRight_299_299_3_/cpu:0__all	245.27	239.88
*benchmarkRandomFlipLeftRight_299_299_3___all	252.71	241.20

* There were no changes made to `RandomFlipLeftRight in this PR

Let me know if you would like me to add more performance tests for the other methods. I don't think there should be any performance impact, but I'm happy to add more if you'd like.

[JoshVarty]

/cc @martinwicke @jhseu

I would still like to include support for the random flip ops as well, but I'll do that in another PR if this looks acceptable.

[rmlarsen]

@martinwicke could you take a look, please?

[tensorflowbutler]

Nagging Assignee: It has been 14 days with no activity and this issue has an assignee. Please update the label and/or status accordingly.

[martinwicke]

I'm sorry this took me so long. Could you fix the conflicts? I think it's just that proper name_scopes were added to all functions in this module, which is nice and should be easy to merge.

[JoshVarty]

Introducing two methods for _rot90_3D and _rot90_4D is unfortunate. I'd definitely welcome a better suggestion. The other alternative was to stagger if shape.ndims == 3 checks throughout the method which didn't really help readability either.

[martinwicke]

I think this is fine.

[JoshVarty]

@martinwicke I've merged the commits and I think I've done so properly. All the tests pass and there are still no performance regressions so it's probably ready for a code review.

I also introduced _AssertAtLeast3DImage() in the style of _Assert3DImage() which had been added since I originally submitted this PR.

[martinwicke]

Sanity complains about indentation:

FAIL: Found 5 non-whitelited pylint errors:
tensorflow/python/ops/image_ops_impl.py:422: [C0330(bad-continuation), ] Wrong continued indentation (add 2 spaces).

tensorflow/python/ops/image_ops_impl.py:428: [C0330(bad-continuation), ] Wrong continued indentation (add 3 spaces).

tensorflow/python/ops/image_ops_impl.py:437: [C0301(line-too-long), ] Line too long (82/80)

tensorflow/python/ops/image_ops_impl.py:458: [C0330(bad-continuation), ] Wrong continued indentation (add 3 spaces).

tensorflow/python/ops/image_ops_test.py:1125: [C0330(bad-continuation), ] Wrong continued indentation (add 22 spaces).

[googlebot]

So there's good news and bad news.

👍 The good news is that everyone that needs to sign a CLA (the pull request submitter and all commit authors) have done so. Everything is all good there.

😕 The bad news is that it appears that one or more commits were authored by someone other than the pull request submitter. We need to confirm that all authors are ok with their commits being contributed to this project. Please have them confirm that here in the pull request.

Note to project maintainer: This is a terminal state, meaning the cla/google commit status will not change from this State. It's up to you to confirm consent of the commit author(s) and merge this pull request when appropriate.

[JoshVarty]

Sorry for the delay fixing indentation. I forgot to run the sanity checks locally. :(

Thanks for taking care of it!