提交 · 14b2d686d68696f90dbd08564b11af04066ce291 · xxadev / tensorflow

07 7月, 2020 1 次提交
- S
  Shift forward-compat window for NMS padded to fix TFLite conversion in 2.3 · 1725ab69
  由 Sachin Joglekar 提交于 7月 06, 2020
```
PiperOrigin-RevId: 319856402
Change-Id: I60eeb4e411609511046cc4676c2f3ad630b3769d
```
  1725ab69
23 6月, 2020 1 次提交
- A
  Enable tf.image.ssim to run on tensor-like inputs and add SSIM name space. · 834f2bd7
  由 A. Unique TensorFlower 提交于 6月 22, 2020
```
PiperOrigin-RevId: 317676190
Change-Id: I08408eaf397ace235b5f50513096cbd9ba46d5a8
```
  834f2bd7
06 6月, 2020 1 次提交
- V
  Add more error information about the current shape that isn't correct. · 86c745a1
  由 Vijay Vasudevan 提交于 6月 05, 2020
```
PiperOrigin-RevId: 314962826
Change-Id: Ie45124beb74238bc683f45f4438df21247b299ca
```
  86c745a1
04 6月, 2020 1 次提交
- A
  Fix formatting in image_flip_left_right · 4033415a
  由 A. Unique TensorFlower 提交于 6月 03, 2020
```
PiperOrigin-RevId: 314550809
Change-Id: I2565526c7c8ad2463dcaa482bd048778bee5d54c
```
  4033415a
16 5月, 2020 1 次提交
- E
  Add dispatch support to more Python APIs. · 77245d07
  由 Edward Loper 提交于 5月 15, 2020
```
PiperOrigin-RevId: 311763060
Change-Id: Ib35371483aa083e245996508a82fd13d8ac43131
```
  77245d07
15 5月, 2020 1 次提交

Addresses https://github.com/tensorflow/tensorflow/issues/38694 by ensuring... · 24c75ce5

由 Hye Soo Yang 提交于 5月 14, 2020

Addresses https://github.com/tensorflow/tensorflow/issues/38694 by ensuring TensorShape v1/v2 backwards compatibility.

PiperOrigin-RevId: 311670326
Change-Id: I0e7045ff4eb19cb0096d6fa41a494c8f9a6b85c5

24c75ce5

13 5月, 2020 6 次提交
- Y
  Update doc example of v1 to keep old behavior with usage of tf.compat.v1.image.extract_glimpse · c00af599
  由 Yong Tang 提交于 5月 06, 2020
```
Signed-off-by: NYong Tang <yong.tang.github@outlook.com>
```
  c00af599
- Y
  Reroute tf.compat.v1.extract_glimpse to use gen_image_ops.extract_glimpse (old API) · 23890c00
  由 Yong Tang 提交于 4月 28, 2020
```
This fix reroute tf.compat.v1.extract_glimpse to use gen_image_ops.extract_glimpse,
so that the behavior of TF 1.x remains the same.
Signed-off-by: NYong Tang <yong.tang.github@outlook.com>
```
  23890c00
- Y
  Expand test case to cover both old kernel (ExtractGlimpse) and new kernel (ExtractGlimpseV2) · 9b84edeb
  由 Yong Tang 提交于 4月 27, 2020
```
Signed-off-by: NYong Tang <yong.tang.github@outlook.com>
```
  9b84edeb
- Y
  Use ExtractGlimpseV2 and ExtractGlimpse to make sure C++ kernel is backward compatible · 756b7ed2
  由 Yong Tang 提交于 4月 27, 2020
```
Signed-off-by: NYong Tang <yong.tang.github@outlook.com>
```
  756b7ed2
- Y
  Fix incorrect doc example with centered=False · 353d22eb
  由 Yong Tang 提交于 4月 15, 2020
```
Signed-off-by: NYong Tang <yong.tang.github@outlook.com>
```
  353d22eb
- Y
  Fix incorrect doc test · 3fc74213
  由 Yong Tang 提交于 4月 15, 2020
```
Signed-off-by: NYong Tang <yong.tang.github@outlook.com>
```
  3fc74213
28 4月, 2020 1 次提交
- A
  Saturate when converting image between datatypes in adjust_jpeg_quality. · 5cac87ab
  由 A. Unique TensorFlower 提交于 4月 27, 2020
```
PiperOrigin-RevId: 308677540
Change-Id: Ibd1f2cfebb11fce2dcce8cafb1c7e191ce8ae368
```
  5cac87ab
02 4月, 2020 1 次提交
- J
  Correct forward_compatibility behavior. · 49690e53
  由 Jinliang Wei 提交于 4月 01, 2020
```
PiperOrigin-RevId: 304199164
Change-Id: I25b21a7b3f1f82bdd23b56eed0cd451f05897907
```
  49690e53
31 3月, 2020 1 次提交

Fixed a few bugs in non_max_suppression_padded. · ad692522

由 Jinliang Wei 提交于 3月 30, 2020

Properly deal with boxes that are suppressed by score_threshold when canonicalizing coordinates.

Also suppress boxes whose score equals to the score_threshold to be compatible with previous implementation's behavior.

Computing `num_boxes_after_padding // tile_size` is moved outside of loop condition to make it easier for Grappler's constant folding optimization.

Output is always padded to max_output_size even when the input size (number of boxes) is less than max_output_size.

Properly handle inputs with dynamic shapes.

Due to issues with TFLite conversion, we set a forward compatibility window of 3 weeks. Until 2020/4/20, calling non_max_suppression_padded defaults to the old implementation.

PiperOrigin-RevId: 303791501
Change-Id: I7307634bf493e0fb971b50e71548fd97e27e2073

ad692522

27 3月, 2020 1 次提交
- A
  Clarify that ResizeMethod.BILINEAR is actually tf.image.ResizeMethod.BILINEAR, · c94f1f00
  由 A. Unique TensorFlower 提交于 3月 26, 2020
```
etc.

PiperOrigin-RevId: 303201845
Change-Id: I59b3e0b02ebe00264d92a88231f628b1c365f51c
```
  c94f1f00
25 3月, 2020 1 次提交

Replace non_max_suppression_padded with a new implementation that supports... · 448a04e7

由 A. Unique TensorFlower 提交于 3月 24, 2020

Replace non_max_suppression_padded with a new implementation that supports batched inputs. This new implementation is considerably faster than previous implementations on TPU and GPU.

PiperOrigin-RevId: 302730463
Change-Id: I3ff9a1204dc892b7c6688b6b400281ff9cd5331f

448a04e7

18 3月, 2020 1 次提交
- Y
  Remove the failing test on macos. · a12180b8
  由 Yash Katariya 提交于 3月 17, 2020
```
PiperOrigin-RevId: 301424820
Change-Id: Ib0623bab8459c758473e16b6c4be2751ab0f7577
```
  a12180b8
15 3月, 2020 1 次提交
- A
  
  Changed example and output · afab9ac5
  由 Ashutosh Hathidara 提交于 3月 14, 2020
  
  afab9ac5
14 3月, 2020 1 次提交
- A
  
  Added preprocessing and corrected example of RGB images before feeding rgb_to_yub · 1aab9954
  由 Ashutosh Hathidara 提交于 3月 14, 2020
  
  1aab9954
13 3月, 2020 1 次提交
- A
  
  Added code for preprocessing of YUB images before feeding yub_to_rgb · 52169bdc
  由 Ashutosh Hathidara 提交于 3月 13, 2020
  
  52169bdc
05 3月, 2020 1 次提交
- V
  
  Export encode_png to tf.io · 4b2c0d0a
  由 vijayphoenix 提交于 3月 05, 2020
  
  4b2c0d0a
03 3月, 2020 2 次提交
- “
  
  Minor additional fix to bilinear markdown · c9a80ddb
  由 “jaketae” 提交于 3月 03, 2020
  
  c9a80ddb
- “
  
  Fixed inline code markdown for in docstring · fb600b58
  由 “jaketae” 提交于 3月 03, 2020
  
  fb600b58
27 2月, 2020 1 次提交
- F
  
  fix resize image bug · e571da0e
  由 fsx950223 提交于 2月 27, 2020
  
  e571da0e
11 2月, 2020 1 次提交
- K
  
  minor spelling tweaks · bd8e308b
  由 Kazuaki Ishizaki 提交于 12月 10, 2019
  
  bd8e308b
06 2月, 2020 1 次提交
- M
  Add examples to image.resize and rot90 docstring. · b7f8c170
  由 Mark Daoust 提交于 2月 05, 2020
```
PiperOrigin-RevId: 293403075
Change-Id: If8484c43ee7388340532e84c72a3a5d92762f786
```
  b7f8c170
24 1月, 2020 1 次提交
- Y
  Add a newline before the doctest and indent it at the same level. · 3ba8ba55
  由 Yash Katariya 提交于 1月 23, 2020
```
PiperOrigin-RevId: 291182588
Change-Id: Ib56ba891a86f7a60179a0a5497268b0e0e3315ef
```
  3ba8ba55
22 1月, 2020 1 次提交
- R
  
  add numbers again · 8c60c55e
  由 Rick Wierenga 提交于 1月 21, 2020
  
  8c60c55e
21 1月, 2020 2 次提交
- R
  
  fix numpy array formatting · eeef19c2
  由 Rick Wierenga 提交于 1月 21, 2020
  
  eeef19c2
- R
  
  Update image_ops_impl.py · f1c5f136
  由 Rick Wierenga 提交于 1月 21, 2020
  
  f1c5f136
20 1月, 2020 3 次提交
- R
  
  remove empty lines · 60596fa2
  由 Rick Wierenga 提交于 1月 20, 2020
  
  60596fa2
- R
  
  remove redundant 0's · 5cf792e4
  由 Rick Wierenga 提交于 1月 20, 2020
  
  5cf792e4
- R
  
  Add usage example to pad_to_bounding_box · a72b7e26
  由 Rick Wierenga 提交于 1月 20, 2020
  
  a72b7e26
18 1月, 2020 2 次提交
- P
  Fix doc test error in `image_ops_impl.py` · 62f302ec
  由 Pavithra Vijay 提交于 1月 17, 2020
```
PiperOrigin-RevId: 290382477
Change-Id: Ie00c667e3b6c1ef21ff45515e078e132efc01e84
```
  62f302ec
- D
  server_lib.create_local_server() should create TF server with job_name=localhost · 035158d1
  由 Dong Lin 提交于 1月 17, 2020
```
PiperOrigin-RevId: 290361995
Change-Id: I277c80b6f910f1f3bf20d0b26fdad24a95566786
```
  035158d1
08 1月, 2020 1 次提交
- Y
  Add a newline before doctests. · 20a33d8d
  由 Yash Katariya 提交于 1月 07, 2020
```
PiperOrigin-RevId: 288536232
Change-Id: I35b6df99a31fa458edeefd4bc125fffd8fede11a
```
  20a33d8d
07 1月, 2020 3 次提交
- M
  
  Update image_ops_impl.py · 3ed8e55b
  由 Mbah-Javis 提交于 1月 06, 2020
  
  3ed8e55b
- M
  
  Update image_ops_impl.py · d5aab92a
  由 Mbah-Javis 提交于 1月 06, 2020
  
  d5aab92a
- M
  
  Update image_ops_impl.py · e396e081
  由 Mbah-Javis 提交于 1月 06, 2020
  
  e396e081

xxadev / tensorflow 与 Fork 源项目一致

xxadev / tensorflow
与 Fork 源项目一致