- 07 7月, 2020 1 次提交
-
-
由 Sachin Joglekar 提交于
PiperOrigin-RevId: 319856402 Change-Id: I60eeb4e411609511046cc4676c2f3ad630b3769d
-
- 23 6月, 2020 1 次提交
-
-
由 A. Unique TensorFlower 提交于
PiperOrigin-RevId: 317676190 Change-Id: I08408eaf397ace235b5f50513096cbd9ba46d5a8
-
- 06 6月, 2020 1 次提交
-
-
由 Vijay Vasudevan 提交于
PiperOrigin-RevId: 314962826 Change-Id: Ie45124beb74238bc683f45f4438df21247b299ca
-
- 04 6月, 2020 1 次提交
-
-
由 A. Unique TensorFlower 提交于
PiperOrigin-RevId: 314550809 Change-Id: I2565526c7c8ad2463dcaa482bd048778bee5d54c
-
- 16 5月, 2020 1 次提交
-
-
由 Edward Loper 提交于
PiperOrigin-RevId: 311763060 Change-Id: Ib35371483aa083e245996508a82fd13d8ac43131
-
- 15 5月, 2020 1 次提交
-
-
https://github.com/tensorflow/tensorflow/issues/38694由 Hye Soo Yang 提交于
Addresses https://github.com/tensorflow/tensorflow/issues/38694 by ensuring TensorShape v1/v2 backwards compatibility. PiperOrigin-RevId: 311670326 Change-Id: I0e7045ff4eb19cb0096d6fa41a494c8f9a6b85c5
-
- 13 5月, 2020 6 次提交
-
-
由 Yong Tang 提交于
Signed-off-by: NYong Tang <yong.tang.github@outlook.com>
-
由 Yong Tang 提交于
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>
-
由 Yong Tang 提交于
Signed-off-by: NYong Tang <yong.tang.github@outlook.com>
-
由 Yong Tang 提交于
Signed-off-by: NYong Tang <yong.tang.github@outlook.com>
-
由 Yong Tang 提交于
Signed-off-by: NYong Tang <yong.tang.github@outlook.com>
-
由 Yong Tang 提交于
Signed-off-by: NYong Tang <yong.tang.github@outlook.com>
-
- 28 4月, 2020 1 次提交
-
-
由 A. Unique TensorFlower 提交于
PiperOrigin-RevId: 308677540 Change-Id: Ibd1f2cfebb11fce2dcce8cafb1c7e191ce8ae368
-
- 02 4月, 2020 1 次提交
-
-
由 Jinliang Wei 提交于
PiperOrigin-RevId: 304199164 Change-Id: I25b21a7b3f1f82bdd23b56eed0cd451f05897907
-
- 31 3月, 2020 1 次提交
-
-
由 Jinliang Wei 提交于
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
-
- 27 3月, 2020 1 次提交
-
-
由 A. Unique TensorFlower 提交于
etc. PiperOrigin-RevId: 303201845 Change-Id: I59b3e0b02ebe00264d92a88231f628b1c365f51c
-
- 25 3月, 2020 1 次提交
-
-
由 A. Unique TensorFlower 提交于
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
-
- 18 3月, 2020 1 次提交
-
-
由 Yash Katariya 提交于
PiperOrigin-RevId: 301424820 Change-Id: Ib0623bab8459c758473e16b6c4be2751ab0f7577
-
- 15 3月, 2020 1 次提交
-
-
由 Ashutosh Hathidara 提交于
-
- 14 3月, 2020 1 次提交
-
-
由 Ashutosh Hathidara 提交于
-
- 13 3月, 2020 1 次提交
-
-
由 Ashutosh Hathidara 提交于
-
- 05 3月, 2020 1 次提交
-
-
由 vijayphoenix 提交于
-
- 03 3月, 2020 2 次提交
- 27 2月, 2020 1 次提交
-
-
由 fsx950223 提交于
-
- 11 2月, 2020 1 次提交
-
-
由 Kazuaki Ishizaki 提交于
-
- 06 2月, 2020 1 次提交
-
-
由 Mark Daoust 提交于
PiperOrigin-RevId: 293403075 Change-Id: If8484c43ee7388340532e84c72a3a5d92762f786
-
- 24 1月, 2020 1 次提交
-
-
由 Yash Katariya 提交于
PiperOrigin-RevId: 291182588 Change-Id: Ib56ba891a86f7a60179a0a5497268b0e0e3315ef
-
- 22 1月, 2020 1 次提交
-
-
由 Rick Wierenga 提交于
-
- 21 1月, 2020 2 次提交
-
-
由 Rick Wierenga 提交于
-
由 Rick Wierenga 提交于
-
- 20 1月, 2020 3 次提交
-
-
由 Rick Wierenga 提交于
-
由 Rick Wierenga 提交于
-
由 Rick Wierenga 提交于
-
- 18 1月, 2020 2 次提交
-
-
由 Pavithra Vijay 提交于
PiperOrigin-RevId: 290382477 Change-Id: Ie00c667e3b6c1ef21ff45515e078e132efc01e84
-
由 Dong Lin 提交于
PiperOrigin-RevId: 290361995 Change-Id: I277c80b6f910f1f3bf20d0b26fdad24a95566786
-
- 08 1月, 2020 1 次提交
-
-
由 Yash Katariya 提交于
PiperOrigin-RevId: 288536232 Change-Id: I35b6df99a31fa458edeefd4bc125fffd8fede11a
-
- 07 1月, 2020 3 次提交
-
-
由 Mbah-Javis 提交于
-
由 Mbah-Javis 提交于
-
由 Mbah-Javis 提交于
-