tidl/tidl-api.git
21 months agoUpdate TIDL-API manifest for version 1.3.1
Yuan Zhao [Tue, 3 Sep 2019 16:37:08 +0000 (11:37 -0500)]
Update TIDL-API manifest for version 1.3.1

21 months agoUpdate squeeze net reference output
Yuan Zhao [Tue, 3 Sep 2019 15:58:18 +0000 (10:58 -0500)]
Update squeeze net reference output

- 6 out of 1000 outputs changed with the newly imported network

21 months agoPLSDK-2986: update TIDL models for mobilenetV1, inceptionNetV1, squeezeNetV1.
Jianzhong Xu [Fri, 30 Aug 2019 21:32:58 +0000 (17:32 -0400)]
PLSDK-2986: update TIDL models for mobilenetV1, inceptionNetV1, squeezeNetV1.

21 months agoAdjust example heap sizes with new TIDL library
Yuan Zhao [Fri, 30 Aug 2019 19:14:31 +0000 (14:14 -0500)]
Adjust example heap sizes with new TIDL library

- The latest TIDL library increased memory requirement slightly.
  Adjust heap sizes in the examples accordingly.
- MCT-1217

21 months agoPrint out imagenet object index in imagenet example
Yuan Zhao [Wed, 28 Aug 2019 21:28:18 +0000 (16:28 -0500)]
Print out imagenet object index in imagenet example

- The object index is helpful information, in addition to text label
  that has already been printed out.
- Offset tensorflow model output by 1 to remove backgroup index
- MCT-1216

21 months agoControl heap size and alloc opt using env vars
Yuan Zhao [Tue, 27 Aug 2019 18:56:02 +0000 (13:56 -0500)]
Control heap size and alloc opt using env vars

- TIDL_PARAM_HEAP_SIZE_EVE, TIDL_PARAM_HEAP_SIZE_DSP,
  TIDL_NETWORK_HEAP_SIZE_EVE, TIDL_NETWORK_HEAP_SIZE_DSP,
  TIDL_EXTMEM_ALLOC_OPT_EVE, TIDL_EXTMEM_ALLOC_OPT_DSP
  are provided to overwrite the heap sizes and heap allocation optimization
  level (1 or 2) that are specified by default or by application.
- MCT-1215

21 months agoChange develop version to 1.3.1 for patch release
Yuan Zhao [Mon, 26 Aug 2019 14:56:57 +0000 (09:56 -0500)]
Change develop version to 1.3.1 for patch release

22 months agoUpdate network binary in TIDL-API to new format
Yuan Zhao [Sun, 28 Jul 2019 05:17:44 +0000 (00:17 -0500)]
Update network binary in TIDL-API to new format

- New network format corresponds to the network data structure update,
  where strideOffsetMethod field moved from sTIDL_Network_t to sTIDL_Layer_t.
  Old format is 483364 bytes, new format is 484384 bytes.
- Relates to: commit 49401e64374a4f0999479245dcd01eab38bec304, MCT-1136
- MCT-1203

23 months agoAdd ssd_multibox_fullnet example
Yuan Zhao [Wed, 26 Jun 2019 14:36:54 +0000 (09:36 -0500)]
Add ssd_multibox_fullnet example

- To demonstrate running jdenet/jdetnet_voc on a single core,
  without paritioning the network.  This is useful for situations
  where SoC only has C66x cores but not EVE cores.
- MCT-1202

23 months agoDump dataQ/minValue/maxValue for TIDL trace
Yuan Zhao [Wed, 19 Jun 2019 15:40:22 +0000 (10:40 -0500)]
Dump dataQ/minValue/maxValue for TIDL trace

- MCT-1201

2 years agoFix g++ 8.3.0 compilation error
Yuan Zhao [Thu, 20 Jun 2019 17:01:24 +0000 (12:01 -0500)]
Fix g++ 8.3.0 compilation error

- Fix a syntax allowed in g++ 7.2.1 (PSDK5.3) but not in 8.3.0 (PSDK6.0)
- Make should report error from loop
- MCT-1199

2 years agoAdd imagenet python example
Yuan Zhao [Tue, 11 Jun 2019 14:37:03 +0000 (09:37 -0500)]
Add imagenet python example

- Show how to interface with EO/EOP's input/output buffer in python.
- Show how to use OpenCV to read and transform image,
  and how to process imagenet's output data.
- Fix EOP construction in examples
- MCT-1197

2 years agoFix unique_ptr that holds an allocated array
Yuan Zhao [Mon, 10 Jun 2019 20:12:03 +0000 (15:12 -0500)]
Fix unique_ptr that holds an allocated array

- Customer reported this problem.  unique_ptr that holds an allocated
  array was created as "unique_ptr<char>", which will call "delete"
  at destruction.  However, the array was created with "new char[]".
  The proper way should be "unique_ptr<char[]>", so that "delete []"
  will be called at destrution.
- One minor trace message update so that we know which type of device
  is being dispatched to.
- MCT-1196

2 years agoUse DSP Built-in Kernels in TIDL-API
Yuan Zhao [Tue, 12 Mar 2019 17:34:56 +0000 (12:34 -0500)]
Use DSP Built-in Kernels in TIDL-API

- Replace previously used kernel wrappers
- MCT-1143, MCT-1154

2 years agoMerge tag 'v01.03.00' into develop
Yuan Zhao [Mon, 4 Mar 2019 21:57:04 +0000 (15:57 -0600)]
Merge tag 'v01.03.00' into develop

TIDL-API 01.03.00 for Processor SDK 5.3

2 years agoMerge branch 'release/v01.03.00' v01.03.00
Yuan Zhao [Mon, 4 Mar 2019 21:56:34 +0000 (15:56 -0600)]
Merge branch 'release/v01.03.00'

2 years agoUpdate TIDL network data structure
Yuan Zhao [Thu, 28 Feb 2019 16:41:04 +0000 (10:41 -0600)]
Update TIDL network data structure

- To be in sync with TIDL library and TIDL import utility
- strideOffsetMethod field moved from sTIDL_Network_t to sTIDL_Layer_t
- Add ReadNetworkBinary util that can read both network formats,
  so that TIDL-API can be compatible with both old and new formats
- Update reference network output due to updated TIDL library
- MCT-1136

2 years ago[segmentation] Add video clip autorewind
Djordje Senicic [Mon, 4 Mar 2019 12:43:58 +0000 (07:43 -0500)]
[segmentation] Add video clip autorewind

Signed-off-by: Djordje Senicic <x0157990@ti.com>
2 years ago[segmentation] Add sample clip with traffic scenes (from pixabay)
Djordje Senicic [Mon, 4 Mar 2019 10:53:38 +0000 (05:53 -0500)]
[segmentation] Add sample clip with traffic scenes (from pixabay)

Signed-off-by: Djordje Senicic <x0157990@ti.com>
2 years agoUpdate documentation for TIDL-API 1.3.0
Yuan Zhao [Fri, 1 Mar 2019 22:14:37 +0000 (16:14 -0600)]
Update documentation for TIDL-API 1.3.0

- MCT-1136

2 years agoClean up ssd_multibox changes
Yuan Zhao [Fri, 1 Mar 2019 17:41:34 +0000 (11:41 -0600)]
Clean up ssd_multibox changes

- PLSDK-2597

2 years ago[ssd_multibox] Addressing review comments
Djordje Senicic [Thu, 28 Feb 2019 17:03:28 +0000 (12:03 -0500)]
[ssd_multibox] Addressing review comments

Signed-off-by: Djordje Senicic <x0157990@ti.com>
2 years agoBump develop branch version to 1.4.0
Yuan Zhao [Thu, 28 Feb 2019 16:59:41 +0000 (10:59 -0600)]
Bump develop branch version to 1.4.0

2 years agoPLSDK-2597
Djordje Senicic [Thu, 28 Feb 2019 13:47:43 +0000 (08:47 -0500)]
PLSDK-2597
  - SSD_Multibox: updated to include slider for run-time probability modification
  - SSD_Multibox: skip grabbing frame input multiple times, as real-time would very based on multicore configuration and network complexity
  - SSD_Multibox: resize and central cropping added; instead of showing rectangles in original image, network input is presented
  - Classification: Toydogs configuration added including models

Signed-off-by: Djordje Senicic <x0157990@ti.com>
2 years agoEnable DSP out-of-order execution in TIDL-API
Yuan Zhao [Fri, 28 Dec 2018 18:06:29 +0000 (12:06 -0600)]
Enable DSP out-of-order execution in TIDL-API

- MCT-1108

2 years agoEnable MNIST example on DSP
Yuan Zhao [Wed, 19 Dec 2018 16:53:18 +0000 (10:53 -0600)]
Enable MNIST example on DSP

- It turns out DSP implementation of InnerProduct layer in TIDL library
  requires input size to be multiple of 8, because it is doing
  aligned 8-byte loads.
- Original LeNet network used in the MNIST example has a second InnerProduct
  layer of size 500, which is not a multiple of 8.  Change the size to 504,
  re-train the network, re-import into TIDL format.  Now the MNIST example
  works correctly on DSP as well.
- MCT-1105

2 years agoMerge tag 'v01.02.02' into develop
Ajay Jayaraj [Wed, 12 Dec 2018 14:10:20 +0000 (08:10 -0600)]
Merge tag 'v01.02.02' into develop

2 years agoMerge branch 'hotfix/v01.02.02' v01.02.02
Ajay Jayaraj [Wed, 12 Dec 2018 14:09:23 +0000 (08:09 -0600)]
Merge branch 'hotfix/v01.02.02'

2 years agoFix memory leak in classification example
Yuan Zhao [Tue, 11 Dec 2018 14:59:22 +0000 (08:59 -0600)]
Fix memory leak in classification example

- MCT-1101

2 years agoclassification: Modify configuration structure to runFullNet before calling construct...
Djordje Senicic [Fri, 7 Dec 2018 12:40:36 +0000 (07:40 -0500)]
classification: Modify configuration structure to runFullNet before calling constructor for Execution Object

Signed-off-by: Djordje Senicic <x0157990@ti.com>
2 years agoUpdated patch version to 2 (TIDL API 1.2.2)
Ajay Jayaraj [Fri, 7 Dec 2018 17:35:46 +0000 (11:35 -0600)]
Updated patch version to 2 (TIDL API 1.2.2)

2 years agoMerge tag 'v01.02.01' into develop
Ajay Jayaraj [Thu, 6 Dec 2018 17:36:50 +0000 (11:36 -0600)]
Merge tag 'v01.02.01' into develop

2 years agoMerge branch 'hotfix/v01.02.01' v01.02.01
Ajay Jayaraj [Thu, 6 Dec 2018 17:35:48 +0000 (11:35 -0600)]
Merge branch 'hotfix/v01.02.01'

2 years agotidl-viewer: Remove executor.h inclusion in utils.cpp
Ajay Jayaraj [Thu, 6 Dec 2018 17:00:35 +0000 (11:00 -0600)]
tidl-viewer: Remove executor.h inclusion in utils.cpp

One of the files in the tidl-viewer build, utils.cpp, was updated to include
executor.h.  This header in turn includes a file from OpenCL, custom.h.

The yocto build of tidl-viewer is a native recipe and hence cannot
include opencl recipes as a dependency to obtain custom.h.

This commit updates utils.cpp to remove the include of executor.h and
custom.h.
(MCT-1100)

2 years agoMerge tag 'v01.02.00' into develop
Ajay Jayaraj [Wed, 28 Nov 2018 23:11:04 +0000 (17:11 -0600)]
Merge tag 'v01.02.00' into develop

2 years agoMerge branch 'release/v01.02.00' v01.02.00
Ajay Jayaraj [Wed, 28 Nov 2018 23:09:41 +0000 (17:09 -0600)]
Merge branch 'release/v01.02.00'

2 years agoAdd jdetnet_voc network and make it the default
Yuan Zhao [Wed, 28 Nov 2018 21:25:57 +0000 (15:25 -0600)]
Add jdetnet_voc network and make it the default

- jdetnet_voc is trained with more object categories than original
  jdetnet.  Make jdetnet_voc the default in the ssd example.  User can
  still use command line options to run the original jdetnet network.
- MCT-1091

2 years agoUpdate imagenet example with new softmax output
Yuan Zhao [Tue, 27 Nov 2018 23:36:44 +0000 (17:36 -0600)]
Update imagenet example with new softmax output

- MCT-1089

2 years agoAdded Python variant of mnist example
Ajay Jayaraj [Tue, 27 Nov 2018 20:56:19 +0000 (14:56 -0600)]
Added Python variant of mnist example

Also fix one_eo_per_frame.py to avoid creating an EVE executor if there are
no EVEs available.
(MCT-1088)

2 years agoAdd link in changelog for mnist example
Yuan Zhao [Tue, 27 Nov 2018 19:30:41 +0000 (13:30 -0600)]
Add link in changelog for mnist example

- MCT-1083

2 years agoAdd MNIST LeNet network model and test input
Yuan Zhao [Tue, 27 Nov 2018 15:58:46 +0000 (09:58 -0600)]
Add MNIST LeNet network model and test input

- Constrained to EVE only for now.
- Add documentation for mnist example.
- MCT-1083

2 years agoAdd mnist example with low compute
Yuan Zhao [Wed, 7 Nov 2018 20:50:59 +0000 (14:50 -0600)]
Add mnist example with low compute

- Show that TIDL API with multiple contexts and pipelined computation
  offers low overhead for small networks as well.
- MCT-1083

2 years agoUpdate reference output for unit tests
Ajay Jayaraj [Mon, 26 Nov 2018 22:28:05 +0000 (16:28 -0600)]
Update reference output for unit tests

A defect fix in the softmax layer necessitated updates to the reference
output of networks using the softmax layer.

(MCT-1087)

2 years agoDocumentation - update 'Using the API' chapter
Ajay Jayaraj [Thu, 15 Nov 2018 22:33:55 +0000 (16:33 -0600)]
Documentation - update 'Using the API' chapter

(MCT-1086)

2 years agoInitialize EO::current_frame_idx_m in constructor
Ajay Jayaraj [Thu, 15 Nov 2018 16:50:37 +0000 (10:50 -0600)]
Initialize EO::current_frame_idx_m in constructor

Initialize ExecutionObject::current_frame_idx_m array to 0 in the
ExecutionObject constructor to prevent out of range entries when
recording trace data.

In a pipelined processing loop, the application executes
ExecutionObject::ProcessFrameWait() on the first frame before it calls
ExecutionObject::ProcessFrameStartAsync. The side effect is that the
current_frame_idx_m is not initialized. This can result in negative
frame indices when writing trace data using ReportTrace or UpdateTrace
leading to memory errors.

Setting ExecutionObject::current_frame_idx_m to 0 in the constructor
avoids this scenario.

(MCT-1085)

2 years agoUpdated version on develop to 1.3.0
Ajay Jayaraj [Wed, 14 Nov 2018 17:18:11 +0000 (11:18 -0600)]
Updated version on develop to 1.3.0

2 years agoUpdated parameter description in doxygen comments
Ajay Jayaraj [Tue, 13 Nov 2018 21:46:30 +0000 (15:46 -0600)]
Updated parameter description in doxygen comments

(MCT-1084)

2 years agoBuild Python bindings library by default
Ajay Jayaraj [Wed, 7 Nov 2018 16:16:04 +0000 (10:16 -0600)]
Build Python bindings library by default

(MCT-1069)

2 years agoUpdated manifest for v1.2.0
Ajay Jayaraj [Thu, 1 Nov 2018 19:00:38 +0000 (14:00 -0500)]
Updated manifest for v1.2.0

(MCT-1060)

2 years agoAdd option to specify object classes list file
Yuan Zhao [Wed, 31 Oct 2018 21:53:30 +0000 (16:53 -0500)]
Add option to specify object classes list file

- so that user can specify a different object classes list file
  without re-compiling the application.
- MCT-1081

2 years agomcbench: image preprocessing, handle layergroups=1
Djordje Senicic [Wed, 31 Oct 2018 17:18:00 +0000 (13:18 -0400)]
mcbench: image preprocessing, handle layergroups=1

- Add image preprocessing for types 1 and 2.
- if layer groups is 1, force all layers to be in the same group

(MCT-1075)

2 years agoclassification: Support different network models
Djordje Senicic [Wed, 31 Oct 2018 17:14:34 +0000 (13:14 -0400)]
classification: Support different network models

jacinto11, mobilenet and inceptionet models can be used with this example.

2 years agoinfer: Add configuration files for inceptionnet and mobilenet that can run as two...
Djordje Senicic [Wed, 31 Oct 2018 17:12:00 +0000 (13:12 -0400)]
infer: Add configuration files for inceptionnet and mobilenet that can run as two layer groups

Signed-off-by: Djordje Senicic <x0157990@ti.com>
2 years agotidl_models: Add mobilenet and inceptionnet models, trained on ImageNet
Djordje Senicic [Wed, 31 Oct 2018 17:10:46 +0000 (13:10 -0400)]
tidl_models: Add mobilenet and inceptionnet models, trained on ImageNet

Signed-off-by: Djordje Senicic <x0157990@ti.com>
2 years agomcbench: Multicore benchmark with minimal overhead
Djordje Senicic [Wed, 24 Oct 2018 06:07:41 +0000 (02:07 -0400)]
mcbench: Multicore benchmark with minimal overhead

- Add required models, input test vectors and platform specific scripts
- Add inference configuration files for multicore benchmarking
- Rename input files to indicate multiple frames and add more inference
  configurations, covered in regression scripts

(MCT-1075)

2 years agoexamples: Add layers group command line parameter
Djordje Senicic [Wed, 24 Oct 2018 06:06:42 +0000 (02:06 -0400)]
examples: Add layers group command line parameter

Signed-off-by: Djordje Senicic <x0157990@ti.com>
2 years agoUpdated Python bindings to reflect API changes
Ajay Jayaraj [Wed, 31 Oct 2018 14:56:26 +0000 (09:56 -0500)]
Updated Python bindings to reflect API changes

(MCT-1080)

2 years agoAdd contextSize: hide OCL_TIDL_CACHE_ALIGN on host
Yuan Zhao [Tue, 30 Oct 2018 15:33:29 +0000 (10:33 -0500)]
Add contextSize: hide OCL_TIDL_CACHE_ALIGN on host

- MCT-1059

2 years agoAdded graph viewer for TIDL API timestamp data
Ajay Jayaraj [Wed, 24 Oct 2018 14:40:08 +0000 (09:40 -0500)]
Added graph viewer for TIDL API timestamp data

1. Created a python matplotlib based utility for viewing timestamp data
   generated from the TIDL API (viewer/execution_graph.py)
2. Minor updates to API internals to add ExecutionObject type, index to
   timestamp output

(MCT-1073)

2 years agoSimplify API for multiple contexts
Ajay Jayaraj [Mon, 15 Oct 2018 16:00:46 +0000 (11:00 -0500)]
Simplify API for multiple contexts

1.  Simplify context API in ExecutionObject. Replace context_id variants to
multiple existing APIs with these two APIs:

        bool AcquireAndRunContext(uint32_t& context_idx,
                                  int frame_idx,
                                  const IODeviceArgInfo& in,
                                  const IODeviceArgInfo& out);

        bool WaitAndReleaseContext(uint32_t  context_idx);

2. The timing methods for host execution in EOPs and EOs:

* GetProcessTimeInMilliSeconds()
* GetHostProcessTimeInMilliSeconds()

are no longer accurate with multiple contexts and pipelining.

Replace these methods and replace with a generic timestamp
based approach. There is a single API call to enable time stamps in an
application:

//! Enable time stamp generation for TIDL API events
bool EnableTimeStamps(const std::string& file = "timestamp.log", size_t
num_frames=32);
If this method is called before TIDL API frame processing, the API will
generate timestamps for events corresponding to each frame (e.g.
EOP::ProcessFrameStartAsync, EOP::ProcessFrameWait, etc.). These
timestamps are then written to file when the user's application
completes.

A separate script is used for post-processing the time stamps and
generating data for the user.

(MCT-1073, MCT-1074)

2 years agoEnqueue multiple frames at device side
Yuan Zhao [Thu, 4 Oct 2018 16:39:08 +0000 (11:39 -0500)]
Enqueue multiple frames at device side

- Previous implementation won't send/enqueue next frame to device
  until the host has received completion message for current frame.
  The improvement is to create multiple sets/contexts of internal
  TIDL input/output buffers at device side, and to send/enqueue next
  frame using a different set/context of internal TIDL input/output
  buffers to device while device is still processing the current frame.
  When device finishes current frame, it can immediately read
  its messageQ and start processing the next frame, without waiting
  for the completion message reaching the host and the hosting sending
  the next frame.
- In pipelined processing of multiple frames, this optimization can
  effectively hide the round-trip communication between host and device.
- Removed deprecated enableInternalInput feature
- MCT-1059

2 years agoQuantization history configuration parameters
Ajay Jayaraj [Mon, 15 Oct 2018 16:49:01 +0000 (11:49 -0500)]
Quantization history configuration parameters

Added the following parameters to Configuration:
* quantHistoryParam1
* quantHistoryParam2
* quantMargin

These parameters can be specified in the configuration file or set
directly in the code.

(MCT-1062)

2 years agoRemoved unused API functionality, added changelog
Ajay Jayaraj [Wed, 3 Oct 2018 21:57:49 +0000 (16:57 -0500)]
Removed unused API functionality, added changelog

(MCT-1062)

2 years agoRefactor imgutils::PreProcImage function
Ajay Jayaraj [Mon, 1 Oct 2018 18:49:16 +0000 (13:49 -0500)]
Refactor imgutils::PreProcImage function

- Renamed to imgutils::PreprocessImage
- Remove alloc/memcpy of data buffer, work off split OpenCV::Mat(s)
- Use Configuration to extract width, height, #channels etc.
- Number of ROIs is always 1 for TIDL API, remove from parameter list

(MCT-1063)

2 years agoAdding pybind11 v2.2.4 to repo
Ajay Jayaraj [Wed, 26 Sep 2018 21:59:49 +0000 (16:59 -0500)]
Adding pybind11 v2.2.4 to repo

- https://github.com/pybind/pybind11.git, branch v2.2
- commit sha: 9a19306fbf30642ca331d0ec88e7da54a96860f9
(MCT-1009)

2 years agoAdded Python 3 bindings for TIDL API
Ajay Jayaraj [Fri, 21 Sep 2018 15:53:30 +0000 (10:53 -0500)]
Added Python 3 bindings for TIDL API

* Using pybind11 v2.2 to add Python 3 bindings to TIDL API classes/methods
  https://pybind11.readthedocs.io/en/stable/index.html
  https://github.com/pybind/pybind11/tree/v2.2
* Leveraging the Python buffer protocol to expose input/output buffers
  from ExecutionObject/ExecutionObjectPipeline to Python application
  code. This eliminates copies between the Python application and the TIDL
  API library. (see examples/pybind/one_eo_per_frame.py).
* Methods renamed to follow Python style guide (PEP8)
* Bindings split across multiple pybind_* source files to reduce compile
  time
* tidl_api/Makefile builds a shared object - tidl.so. Add this so to
  PYTHONPATH to make the tidl module available to the Python interpreter.
  >>> import tidl
  >>> help (tidl)
* See examples/pybind for examples of using the Python bindings

(MCT-1009)

2 years agoRestore version to 1.2.0.0
Yuan Zhao [Wed, 12 Sep 2018 20:05:12 +0000 (15:05 -0500)]
Restore version to 1.2.0.0

2 years agoMerge tag 'v01.01.00.01' into develop
Yuan Zhao [Wed, 12 Sep 2018 20:02:18 +0000 (15:02 -0500)]
Merge tag 'v01.01.00.01' into develop

For PSDK 5.1 release

2 years agoMerge branch 'hotfix/v01.01.00.01' v01.01.00.01
Yuan Zhao [Wed, 12 Sep 2018 20:01:53 +0000 (15:01 -0500)]
Merge branch 'hotfix/v01.01.00.01'

2 years agoOptimize classification perf, report loop avg_fps
Yuan Zhao [Sat, 8 Sep 2018 05:06:48 +0000 (00:06 -0500)]
Optimize classification perf, report loop avg_fps

- Double buffer EOPs to overlap host pre/post-processing
  and device processing.  When EOP contains more than one EO,
  pipeline at EO level rather than at EOP level.
- Compute average FPS across a sliding window of frames
  using host loop iteration/frame time.
- MCT-1049

2 years agoStart hotfix v01.01.00.01
Yuan Zhao [Mon, 10 Sep 2018 15:40:51 +0000 (10:40 -0500)]
Start hotfix v01.01.00.01

2 years agoMerge tag 'v01.01.00.00' into develop
Ajay Jayaraj [Fri, 7 Sep 2018 19:46:02 +0000 (14:46 -0500)]
Merge tag 'v01.01.00.00' into develop

2 years agoMerge branch 'release/v01.01.00.00' v01.01.00.00
Ajay Jayaraj [Fri, 7 Sep 2018 19:45:44 +0000 (14:45 -0500)]
Merge branch 'release/v01.01.00.00'

2 years agoUpdated TIDL API manifest for v1.1
Ajay Jayaraj [Fri, 7 Sep 2018 19:02:12 +0000 (14:02 -0500)]
Updated TIDL API manifest for v1.1

(MCT-1050)

2 years agoexamples:classification: Use configuration.numFrames instead of hard coded big value
Djordje Senicic [Fri, 7 Sep 2018 17:34:37 +0000 (13:34 -0400)]
examples:classification: Use configuration.numFrames instead of hard coded big value

Signed-off-by: Djordje Senicic <x0157990@ti.com>
2 years agoexamples:classification: Increase main window and update picture of TIDL SW stack
Djordje Senicic [Fri, 7 Sep 2018 16:38:47 +0000 (12:38 -0400)]
examples:classification: Increase main window and update picture of TIDL SW stack

- MCT-1049

Signed-off-by: Djordje Senicic <x0157990@ti.com>
2 years agoUpdated version to 1.2.0
Ajay Jayaraj [Thu, 6 Sep 2018 20:34:22 +0000 (15:34 -0500)]
Updated version to 1.2.0

2 years agoClassification example code refactor and clean up
Yuan Zhao [Thu, 6 Sep 2018 16:36:17 +0000 (11:36 -0500)]
Classification example code refactor and clean up

- MCT-1049

2 years agoOptimize examples with EOP double buffering
Yuan Zhao [Wed, 5 Sep 2018 20:04:19 +0000 (15:04 -0500)]
Optimize examples with EOP double buffering

- Improve overall loop performance for imagenet and segmentation
- Update documentation on performance
- MCT-1039

2 years agoUpdates to User's Guide and related examples
Ajay Jayaraj [Wed, 5 Sep 2018 16:56:37 +0000 (11:56 -0500)]
Updates to User's Guide and related examples

Changes:
* Overview chapter, includes a Terminology section.
* Section on different use cases in the "Using the API" chapter.
* Updated the Examples chapter to reflect new examples and AM5749
  benchmarking.
* Added the two_eo_per_frame_opt example to illustrate double buffering.

(MCT-1043)

2 years agoWall cleanup, optimize ssd_multibox
Yuan Zhao [Wed, 5 Sep 2018 17:19:49 +0000 (12:19 -0500)]
Wall cleanup, optimize ssd_multibox

- Fix -Wall errors
- Optimize pipeline execution for ssd_multibox
- MCT-1015

2 years agoVideo input option and document update
Yuan Zhao [Thu, 23 Aug 2018 21:53:58 +0000 (16:53 -0500)]
Video input option and document update

- mp4/avi/mov as pre-recorded video input
- camera as live video input, let user choose video input port #
- refactor examples code
- bookkeep each EO's device/host time inside EOP since EO could be shared
- documentation update on 650MHz EVE
- documentation on video inputs and output
- MCT-1015

2 years agoAdded example to illustrate pipelining across EOs
Ajay Jayaraj [Thu, 30 Aug 2018 19:25:11 +0000 (14:25 -0500)]
Added example to illustrate pipelining across EOs

two_eo_per_frame is a simple example to illustrate using
ExecutionObjectPipeline to split processing a single frame across EVE
and DSP.

(MCT-1048)

2 years agoRefactor examples - test, one_eo_per_frame
Ajay Jayaraj [Mon, 27 Aug 2018 21:41:25 +0000 (16:41 -0500)]
Refactor examples - test, one_eo_per_frame

- Remove code duplication across test/main.cpp,
  test/multiple_executors.cpp and one_eo_per_frame/main.cpp
- Moved common code into common/utils.h, common/utils.cpp

(MCT-1047)

2 years agoclassification - fixed zero size image clip
Ajay Jayaraj [Wed, 29 Aug 2018 13:15:49 +0000 (08:15 -0500)]
classification - fixed zero size image clip

(PLSDK-2250)

2 years agoclassification - adjust example for API updates
Djordje Senicic [Tue, 28 Aug 2018 19:59:40 +0000 (14:59 -0500)]
classification - adjust example for API updates

- Enable operation up to 36fps on AM5749 with EVEs at 650MHz
- DisplayHelp update for number of cores

(PLSDK-2250)

2 years agoAdded an example to illustrate 1 EO per frame
Ajay Jayaraj [Fri, 24 Aug 2018 22:01:36 +0000 (17:01 -0500)]
Added an example to illustrate 1 EO per frame

(MCT-1043)

2 years agoDocumentation - refactoring and updates
Ajay Jayaraj [Wed, 22 Aug 2018 16:25:21 +0000 (11:25 -0500)]
Documentation - refactoring and updates

Changes:
- Added a release notes section with notes for v1.0 and v1.1
- Reworked the intro section
- Added an overview section, changed the API software picture to show
  more detail
- Removed duplicate documentation for tidl::Configuration in the rst
  file, moved documentation to doxygen comments in configuration.h
- Moved "building from source" to FAQs

(MCT-1043)

2 years agoUpdate imagenet to take mp4 input
Yuan Zhao [Wed, 22 Aug 2018 14:52:32 +0000 (09:52 -0500)]
Update imagenet to take mp4 input

- clean up command line options
- MCT-1015, MCT-1039

2 years agoReport memory usage when device allocation fails
Ajay Jayaraj [Fri, 17 Aug 2018 20:14:15 +0000 (15:14 -0500)]
Report memory usage when device allocation fails

TIDL API creates 2 device side heaps:
1. Parameter heap
2. Network heap

The sizes of these heaps are specified in the Configuration object, via
PARAM_HEAP_SIZE and NETWORK_HEAP_SIZE.

Existing behavior: If the heaps are not large enough, allocation on the
device triggers an assertion failure with no indication of how large the
heaps need to be for successfull allocation.

To improve the usability of the API, provide feedback to the user on the
heap sizes required to satisfy device side allocations when any
allocation fails.

Also added `-Wall -Werror` when building examples and fixed failures.

(MCT-1035)

2 years agoExecutionObjectPipeline for executing layersGroups
Yuan Zhao [Fri, 10 Aug 2018 04:42:42 +0000 (23:42 -0500)]
ExecutionObjectPipeline for executing layersGroups

- Add top level ExecutionObjectPipeline class to execute multiple
  layersGroups.
- An ExecutionObjectPipeline is constructed from multiple
  ExecutionObjects, each ExecutionObject executes one layersGroup
  in the network, together they execute consecutive layersGroups.
- Same look and feel as ExecutionObject, e.g. ProcessFrameStartAsync,
  ProcessFrameWait, GetInputBufferPointer, GetOutputBufferPointer
- MCT-1017, MCT-1029

2 years agoModified IODeviceArgInfo to enable pipelining EOs
Ajay Jayaraj [Mon, 13 Aug 2018 19:56:24 +0000 (14:56 -0500)]
Modified IODeviceArgInfo to enable pipelining EOs

(MCT-1030)

2 years agoRemove implementation details from ArgInfo
Ajay Jayaraj [Mon, 13 Aug 2018 17:30:40 +0000 (12:30 -0500)]
Remove implementation details from ArgInfo

Implementation details such as argument kind and PipeInfo should not be
a part of the user facing ArgInfo class. Also, PipeInfo is relevant only
for input/output arguments.

Moved implementation details out of ArgInfo and created 2 new classes:
DeviceArgInfo and IODeviceArgInfo.

DeviceArgInfo inherits from ArgInfo and adds an
argument kind (buffer, local or scalar). IODeviceArgInfo consists of
DeviceArgInfo and PipeInfo.

(MCT-1030)

2 years agoRemove input, output buffers from process kernel
Ajay Jayaraj [Fri, 10 Aug 2018 22:39:22 +0000 (17:39 -0500)]
Remove input, output buffers from process kernel

Input and output OpenCL buffers do not have to be passed into the process
kernel. The host will directly update input and output in the buffers
allocated by the TIDL library via the HostWriteNetInput and
HostReadNetOutput methods.

(MCT-1030)

2 years agoclassification: Update static images, synthetic video clip
Djordje Senicic [Thu, 2 Aug 2018 11:11:42 +0000 (07:11 -0400)]
classification: Update static images, synthetic video clip

(MCT-1031)
Signed-off-by: Djordje Senicic <x0157990@ti.com>
2 years agoSupport layer group id assignment updates
Ajay Jayaraj [Thu, 9 Aug 2018 18:33:58 +0000 (13:33 -0500)]
Support layer group id assignment updates

Provide API support for updating layer -> layer group id assignments
before executing network.

(MCT-1028)

2 years agoAdd support for accessing layer output data
Ajay Jayaraj [Fri, 3 Aug 2018 19:26:14 +0000 (14:26 -0500)]
Add support for accessing layer output data

* Added API methods to enable tracing and dump layer output buffers to
  file
  1. ExecutionObject::WriteLayerOutputsToFile
  2. ExecutionObject::GetOutputsFromAllLayers
  3. ExecutionObject::GetOutputFromLayer
  See examples/layer_output/main.cpp, ProcessTrace() for examples of
  using these tracing APIs.
* Added a 'enableTrace' field to the configuration file. Set to true
  to enable tracing.
* Update configuration file parser to allow comments, blank lines

Note:
Tracing requires ti-opencl, SHA-ID 188cc38627315d3041b80ee86e6ca33087348062
or newer.

(MCT-1023)

2 years agoMerge tag 'v01.00.00.03' into develop
Yuan Zhao [Wed, 25 Jul 2018 15:20:43 +0000 (10:20 -0500)]
Merge tag 'v01.00.00.03' into develop

Hotfix v01.00.00.03 for PSDK 5.0

2 years agoMerge branch 'hotfix/v01.00.00.03' v01.00.00.03
Yuan Zhao [Wed, 25 Jul 2018 15:19:50 +0000 (10:19 -0500)]
Merge branch 'hotfix/v01.00.00.03'

2 years agoUpdate to version 01.00.00.03
Yuan Zhao [Wed, 25 Jul 2018 15:11:07 +0000 (10:11 -0500)]
Update to version 01.00.00.03