summary |
shortlog | log |
commit |
commitdiff |
tree
first ⋅ prev ⋅ next
Inki Dae [Fri, 21 Aug 2020 00:19:52 +0000 (09:19 +0900)]
Convert Signed64 type to internal one correctly
Change-Id: I789c0483f65262bc7cba896cc2eee0382cb480f8
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Mon, 10 Aug 2020 02:46:36 +0000 (11:46 +0900)]
Consider two and more models inference support
This patch considers two and more models inference.
Mediavision has one backend engine context per a model.
However, ARMNN requires one runtime engine for all models.
So this patch makes it to use a same runtime instance
for two and more models.
Change-Id: Icc2a42fe20fefbc0b76806e7d107b5ad961bded0
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Mon, 3 Aug 2020 09:36:32 +0000 (18:36 +0900)]
Add Signed64 tensor data type support
Change-Id: I30d70142185158d1eee5f2031e02828124dfd9ce
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Hyunsoo Park [Tue, 7 Jul 2020 08:10:43 +0000 (17:10 +0900)]
Add init code for member variable
Change-Id: I5906254f4327ce05933f4faf99aec2c96d6151d4
Signed-off-by: Hyunsoo Park <hance.park@samsung.com>
Inki Dae [Thu, 4 Jun 2020 07:56:35 +0000 (16:56 +0900)]
Fix initializer list coding rule
Change-Id: I40a673c88db39892135e76f5685e0dc333a53eb0
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Thu, 4 Jun 2020 05:38:01 +0000 (14:38 +0900)]
Fix coding style based on Tizen SE C++ Coding Rule
Tizen SE C++ Coding Rule:
https://code.sec.samsung.net/confluence/pages/viewpage.action?pageId=
160925159
Change-Id: Icbc6dc69923927758bf734bd0b348c2bfa5729f0
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Wed, 3 Jun 2020 09:05:27 +0000 (18:05 +0900)]
Add SetPrivateData function
This function is needed for inference engine interface to pass
armnn specific data to this backend before loading a model file.
Change-Id: I162967bcf0bc7ddb6735a85c0381420183b4830b
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Fri, 15 May 2020 01:49:57 +0000 (10:49 +0900)]
packaging: build only on arm and aarch64
ARMNN is NN runtime library for ARM system so do not build
it on other architecture.
Change-Id: I5cc692abe635b331c7efae9f912b94ea47f2e035
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Fri, 15 May 2020 01:33:28 +0000 (10:33 +0900)]
fix build error on aarch64
In aarch64, size_t is 64bit data and so make sure to use correct
format specifier, and minor cleanup.
Change-Id: I30e3147aaba6e4a9cc1dc87f1fd540054a9a204f
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Fri, 17 Apr 2020 07:47:45 +0000 (16:47 +0900)]
change enumeration values to new ones
some enumeration values of inference-engine-interface have been
updated so change them.
Change-Id: Ia057060061027647873fde09911a0f17fb12e830
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Thu, 9 Apr 2020 06:06:11 +0000 (15:06 +0900)]
packaging: add armnn build requirement
Change-Id: I04c4116253ee5576eb96df4edf1bf98af943c0f6
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Mon, 24 Feb 2020 05:47:07 +0000 (14:47 +0900)]
Drop GetInferenceResult function
Change-Id: Iaecaebe2eb1219a4903c06ff06dde3eb7e69e683
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Mon, 24 Feb 2020 05:17:05 +0000 (14:17 +0900)]
Add more logs and fallback device
Change-Id: Ia3f50b639790edbd47ffb7ec4edc42eb3090f4c2
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Thu, 20 Feb 2020 01:39:40 +0000 (10:39 +0900)]
Fix typo
Change-Id: I7aa4e6733f1743bbe6d3ac1a3f65468e6ad78c0e
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Wed, 19 Feb 2020 08:01:56 +0000 (17:01 +0900)]
Use ARMNN internal data type instead of specific one
Change-Id: I75393f94465b2d3638f66ccc2207cb60655751db
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Tue, 18 Feb 2020 04:20:01 +0000 (13:20 +0900)]
Drop dead functions
AllocateTensorBuffer and ReleaseTenosrBuffer functions aren't needed
anymore so drop them.
Change-Id: I0d3f0a7080bb3c668e903d8116569dca811e652a
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Tue, 18 Feb 2020 04:17:01 +0000 (13:17 +0900)]
Fix build warnings
Change-Id: I64bfaed5fb6b197835065ac9e0715c6d9ffd8b39
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Tue, 18 Feb 2020 04:12:36 +0000 (13:12 +0900)]
Consider various NN model formats
As of now, only TFLite model is supported, and other will be
supported layer.
Change-Id: Ie255875032bfc6051da7a14913564bb9548c69b7
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Tue, 18 Feb 2020 02:37:41 +0000 (11:37 +0900)]
Add multiple input and output tensors support
Change-Id: Ibde983e23f3ba208b29494ece236675213aa52ef
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Tue, 18 Feb 2020 01:56:36 +0000 (10:56 +0900)]
Add user desired input and output layer support
Change-Id: I33216ac409f8087d166154ff57b1cec4b2a80879
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Fri, 14 Feb 2020 04:24:09 +0000 (13:24 +0900)]
Receive output tensor buffers for the inference
This patch receives tensor buffers allocated by upper layer
and makes them to be used for the inference.
As for this, it implements GetOutputTensorBuffers callback.
Change-Id: Ic8a3978205b38de0b1b93cd61b6fec4e98c62f3a
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Thu, 13 Feb 2020 08:34:09 +0000 (17:34 +0900)]
Rename input/output binding members
Change-Id: I8fc122e1eb5c42bdc444930b40015884b5080648
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Thu, 13 Feb 2020 07:32:33 +0000 (16:32 +0900)]
Use input tensor buffers coming from upper layer
This patch makes armnn backend to use input tensor buffers
coming from Inference layer for the inference.
As for this, it addes new several callbacks, and drops unnecessary ones.
Change-Id: Id922a1ecef70e36e781c66f363995a261713b6b0
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Wed, 12 Feb 2020 04:33:45 +0000 (13:33 +0900)]
Use a generic channel value
Nowi, Inference layer will pass a generic channel value
instead of OpenCV specific.
Change-Id: Ibfabe6d3d93cb72cd7ff68479da83cd06f386246
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Tue, 11 Feb 2020 06:29:51 +0000 (15:29 +0900)]
Receive model file format from inference engine interface
Change-Id: I9a63c2d214221eb56049a111d26e1123f7e7c58c
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Mon, 10 Feb 2020 06:28:58 +0000 (15:28 +0900)]
Implement GetBackendCapacity callback
This patch implements GetBackendCapacity callback and
provides device types that ARMNN can run the inference.
Change-Id: If67f53f8e0932ec997dd3ba03bf065dc64056627
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Fri, 7 Feb 2020 06:43:13 +0000 (15:43 +0900)]
Consider one more inference target devices
Change-Id: I51b9a04f60e4b4fe873d8d8a1fda6a122532ce31
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Wed, 5 Feb 2020 09:31:21 +0000 (18:31 +0900)]
Drop SetInputTensorParam/SetOutputTensorParam callbacks
Change-Id: I1ee785034aa06b19fdba260c87f9081a7d312952
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Wed, 5 Feb 2020 08:29:30 +0000 (17:29 +0900)]
Add logs to print out a given target device
Change-Id: Ie85dfbc9e731e0eeceab6fc869c4edfdaea3d3b7
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Wed, 5 Feb 2020 07:18:05 +0000 (16:18 +0900)]
code refactoring to Inference initialization process
This patch uses new Load callback and constructor,
and replaces Init/DeInit callbacks with BindBackend/UnbindBackend callbacks.
Change-Id: Iec12fc15a3754ee312e6352201497fa9f78ac4d5
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Inki Dae [Wed, 22 Jan 2020 09:28:14 +0000 (18:28 +0900)]
add initial code
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Tizen Infrastructure [Wed, 22 Jan 2020 09:10:28 +0000 (09:10 +0000)]
Initial empty repository
projects / platform / core / multimedia / inference-engine-armnn.git / log