MIVisionX 1.3.0 - Release updates (#133)

* Readme Updates - 1.3.0

* New Logo added

* NNIR Bug Fix

README.md

@@ -1,10 +1,9 @@
 [![MIT licensed](https://img.shields.io/badge/license-MIT-blue.svg)](https://opensource.org/licenses/MIT)
 [![Build Status](https://travis-ci.org/GPUOpen-ProfessionalCompute-Libraries/MIVisionX.svg?branch=master)](https://travis-ci.org/GPUOpen-ProfessionalCompute-Libraries/MIVisionX)
 
-
 <p align="center"><img width="60%" src="docs/images/MIVisionX.png" /></p>
 
-MIVisionX toolkit is a comprehensive computer vision and machine intelligence libraries, utilities and applications bundled into a single toolkit. AMD MIVisionX delivers highly optimized open source implementation of the Khronos OpenVX™ and OpenVX™ Extensions along with Convolution Neural Net Model Compiler & Optimizer supporting ONNX, and Khronos NNEF™ exchange formats. The toolkit allows for rapid prototyping and deployment of optimized workloads on a wide range of computer hardware, including small embedded x86 CPUs, APUs, discrete GPUs, and heterogeneous servers.
+MIVisionX toolkit is a set of comprehensive computer vision and machine intelligence libraries, utilities, and applications bundled into a single toolkit. AMD MIVisionX delivers highly optimized open source implementation of the Khronos OpenVX™ and OpenVX™ Extensions along with Convolution Neural Net Model Compiler & Optimizer supporting ONNX, and Khronos NNEF™ exchange formats. The toolkit allows for rapid prototyping and deployment of optimized workloads on a wide range of computer hardware, including small embedded x86 CPUs, APUs, discrete GPUs, and heterogeneous servers.
 
 * [AMD OpenVX](#amd-openvx)
 * [AMD OpenVX Extensions](#amd-openvx-extensions)
@@ -31,7 +30,7 @@ MIVisionX toolkit is a comprehensive computer vision and machine intelligence li
 AMD OpenVX ([amd_openvx](amd_openvx#amd-openvx-amd_openvx)) is a highly optimized open source implementation of the <a href="https://www.khronos.org/openvx/" target="_blank">Khronos OpenVX</a> computer vision specification. It allows for rapid prototyping as well as fast execution on a wide range of computer hardware, including small embedded x86 CPUs and large workstation discrete GPUs.
 
 ## AMD OpenVX Extensions
-The OpenVX framework provides a mechanism to add new vision functions to OpenVX by 3rd party vendors. This project has below OpenVX [modules](amd_openvx_extensions#amd-openvx-extensions-amd_openvx_extensions) and utilities to extend [amd_openvx](amd_openvx#amd-openvx-amd_openvx) project, which contains the AMD OpenVX Core Engine.
+The OpenVX framework provides a mechanism to add new vision functions to OpenVX by 3rd party vendors. This project has below mentioned OpenVX [modules](amd_openvx_extensions#amd-openvx-extensions-amd_openvx_extensions) and utilities to extend [amd_openvx](amd_openvx#amd-openvx-amd_openvx) project, which contains the AMD OpenVX Core Engine.
 
 * [amd_loomsl](amd_openvx_extensions/amd_loomsl): AMD Radeon Loom stitching library for live 360 degree video applications
 * [amd_nn](amd_openvx_extensions/amd_nn#openvx-neural-network-extension-library-vx_nn): OpenVX neural network module
@@ -91,6 +90,8 @@ MIVisionX provides you with tools for accomplishing your tasks throughout the wh
   * install `libprotobuf-dev` and `protobuf-compiler` needed for vx_nn
 * [OpenCV 3.4](https://github.com/opencv/opencv/releases/tag/3.4.0)
   * Set `OpenCV_DIR` environment variable to `OpenCV/build` folder
+* [FFMPEG n4.0.4](https://github.com/FFmpeg/FFmpeg/releases/tag/n4.0.4) - Optional
+  * FFMPEG is required for amd_media & mv_deploy modules
 
 #### Prerequisites setup script for Linux - `MIVisionX-setup.py`
 
@@ -107,9 +108,10 @@ For the convenience of the developer, we here provide the setup script which wil
 ````
 python MIVisionX-setup.py --directory [setup directory - optional]
                           --installer [Package management tool - optional (default:apt-get) [options: Ubuntu:apt-get;CentOS:yum]]
-                          --miopen    [MIOpen Version - optional (default:1.7.1)]      
+                          --miopen    [MIOpen Version - optional (default:1.8.1)]
+                          --ffmpeg    [FFMPEG Installation - optional (default:no) [options:Install ffmpeg - yes]]
 ````
-**Note:** use `--installer yum` for CentOS
+**Note:** use `--installer yum` for **CentOS**
 
 ##### Refer to [Wiki](https://github.com/GPUOpen-ProfessionalCompute-Libraries/MIVisionX/wiki/Suggested-development-workflow) page for developer instructions.
 
@@ -166,9 +168,10 @@ cd MIVisionX
 ````
 python MIVisionX-setup.py --directory [setup directory - optional]
                           --installer [Package management tool - optional (default:apt-get) [options: Ubuntu:apt-get;CentOS:yum]]
-                          --miopen    [MIOpen Version - optional (default:1.7.1)]      
+                          --miopen    [MIOpen Version - optional (default:1.8.1)]
+                          --ffmpeg    [FFMPEG Installation - optional (default:no) [options:Install ffmpeg - yes]]    
 ````
-**Note:** Use `--installer yum` for CentOS
+**Note:** Use `--installer yum` for **CentOS**
 ````
 mkdir build
 cd build
@@ -189,6 +192,7 @@ sudo make install
   * [MIOpen](https://github.com/ROCmSoftwarePlatform/MIOpen) -- make sure to use `-DMIOPEN_BACKEND=OpenCL` option with cmake
 * install [protobuf](https://github.com/protocolbuffers/protobuf/releases/tag/v3.5.2)
 * install [OpenCV](https://github.com/opencv/opencv/releases/tag/3.3.0)
+* install [FFMPEG n4.0.4](https://github.com/FFmpeg/FFmpeg/releases/tag/n4.0.4) - Optional
 * build and install (using `cmake` and `% make install`)
   * executables will be placed in `bin` folder
   * libraries will be placed in `lib` folder
@@ -197,10 +201,6 @@ sudo make install
 * add the installed library path to LD_LIBRARY_PATH environment variable (default `/opt/rocm/mivisionx/lib`)
 * add the installed executable path to PATH environment variable (default `/opt/rocm/mivisionx/bin`)
 
-##### Build `annInferenceApp` using `Qt Creator`
-* build [annInferenceApp.pro](apps/cloud_inference/client_app/annInferenceApp.pro) using Qt Creator
-* or use [annInferenceApp.py](apps/cloud_inference/client_app/annInferenceApp.py) for simple tests
-
 ## Verify the Installation
 
 ### Linux

model_compiler/python/nnir.py

@@ -275,22 +275,20 @@ def updateLocals(self):
         for node in self.nodes:
             for output in node.outputs:
                 count+=1
+                input = self.tensor_dict[node.inputs[0]]
                 if node.type in ['sum', 'add', 'sub', 'mul', 'muladd', 'batch_norm', 'relu', 'leaky_relu', 'softmax']:
-                    input = self.tensor_dict[node.inputs[0]]
                     local = IrTensor()
                     local.setName(output)
                     local.setInfo(input.type, input.shape)
                     local.setFormat(input.format)
                     self.addLocal(local)
                 elif node.type in ['global_avg_pool']:
-                    input = self.tensor_dict[node.inputs[0]]
                     local = IrTensor()
                     local.setName(output)
                     local.setInfo(input.type, [input.shape[0], input.shape[1], 1, 1])
                     local.setFormat(input.format)
                     self.addLocal(local)
                 elif node.type in ['conv', 'avg_pool', 'max_pool', 'lrn']:
-                    input = self.tensor_dict[node.inputs[0]]
                     pads = node.attr.get('pads')
                     strides = node.attr.get('strides')
                     dilations = node.attr.get('dilations')
@@ -315,7 +313,6 @@ def updateLocals(self):
                     local.setFormat(input.format)
                     self.addLocal(local)
                 elif node.type in ['conv_transpose']:
-                    input = self.tensor_dict[node.inputs[0]]
                     pads = node.attr.get('pads')
                     strides = node.attr.get('strides')
                     dilations = node.attr.get('dilations')
@@ -351,7 +348,6 @@ def updateLocals(self):
                     local.setFormat(input.format)
                     self.addLocal(local)
                 elif node.type in ['concat']:
-                    input = self.tensor_dict[node.inputs[0]]
                     axis = node.attr.get('axis')
                     if axis == 1:
                         shape = [input.shape[0], 0, input.shape[2], input.shape[3]]
@@ -373,7 +369,6 @@ def updateLocals(self):
                     local.setFormat(input.format)
                     self.addLocal(local)
                 elif node.type in ['slice']:
-                    input = self.tensor_dict[node.inputs[0]]
                     shape = [input.shape[0], input.shape[1] // len(node.outputs), input.shape[2], input.shape[3]]
                     for name in node.outputs:
                         local = IrTensor()
@@ -382,7 +377,6 @@ def updateLocals(self):
                         local.setFormat(input.format)
                         self.addLocal(local)
                 elif node.type in ['reshape']:
-                    input = self.tensor_dict[node.inputs[0]]
                     param = node.attr.get('shape')
                     if not param:
                         param = self.tensor_dict[node.inputs[1]].shape
@@ -417,7 +411,6 @@ def updateLocals(self):
                     local.setFormat(input.format)
                     self.addLocal(local)
                 elif node.type in ['transpose']:
-                    input = self.tensor_dict[node.inputs[0]]
                     axes = node.attr.get('axes')
                     if axes == [0, 2, 3, 1]:
                         format = 'NHWC'
@@ -433,14 +426,12 @@ def updateLocals(self):
                     local.setFormat(format)
                     self.addLocal(local)
                 elif node.type in ['copy']:
-                    input = self.tensor_dict[node.inputs[0]]
                     local = IrTensor()
                     local.setName(output)
                     local.setInfo(input.type, input.shape)
                     local.setFormat(input.format)
                     self.addLocal(local)
                 elif node.type in ['crop']:
-                    input = self.tensor_dict[node.inputs[0]]
                     reference = self.tensor_dict[node.inputs[1]]
                     axis = node.attr.get('axis')
                     out_shape = []
@@ -455,7 +446,6 @@ def updateLocals(self):
                     local.setFormat(input.format)
                     self.addLocal(local)       
                 elif node.type in ['permute']:
-                    input = self.tensor_dict[node.inputs[0]]    
                     order = node.attr.get("order")   
                     if input.format == 'NCHW' and order == [0, 2, 3, 1]:
                         format = 'NHWC'
@@ -471,7 +461,6 @@ def updateLocals(self):
                     local.setFormat(format)
                     self.addLocal(local)
                 elif node.type in ['flatten']:
-                    input = self.tensor_dict[node.inputs[0]]
                     axis = node.attr.get("axis")
                     if axis == 0:
                         shape = [1, input.shape[0]*input.shape[1]*input.shape[2]*input.shape[3], 1, 1]
@@ -485,7 +474,6 @@ def updateLocals(self):
                     local.setFormat(input.format)
                     self.addLocal(local)
                 elif node.type in ['prior_box']:
-                    input = self.tensor_dict[node.inputs[0]]
                     dim = 1 #for min_size
                     if node.attr.get("max_size") > 0:
                         dim += 1
@@ -501,7 +489,6 @@ def updateLocals(self):
                     local.setFormat(input.format)
                     self.addLocal(local)
                 elif node.type in ['crop_and_resize']:
-                    input = self.tensor_dict[node.inputs[0]]
                     shape = node.attr.get('shape')
                     scaleFactor = node.attr.get('scale')
                     width = shape[0]