| cmake_minimum_required(VERSION 3.22) | |
| project(testOrbbecDevice) | |
| set(CMAKE_CXX_STANDARD 14) | |
| set(OpenCV_DIR "/opencv/build") | |
| find_package(OpenCV REQUIRED) | |
| include_directories( | |
| ${PROJECT_SOURCE_DIR}/ | |
| ${OpenCV_INCLUDE_DIRS} |
| Test on Khadas VIM3(Cortex-A73) (ms) | |||||
|---|---|---|---|---|---|
| Model Name | OpenCV DNN | ONNXRuntime | TFLite + XNNPACK | NCNN | |
| ResNet50 | 418.83 | 351.25 | 230.91 | 186.18 | |
| MobileNet V1 | 62.84 | 75.47 | 31.16 | 42.96 |
OpenCV with TIM-VX is developped to run deep learning models on NPUs (on Khadas VIM3 etc.) using OpenCV with TIM-VX integrated. Code is under review at opencv/opencv#21036. Inference performance on the NPU on Khadas VIM3 can be found in OpenCV Zoo: https://github.com/opencv/opencv_zoo#models--benchmark-results.
We provide two ways compiling OpenCV with TIM-VX backend:
| //#define CV_8U 0 | |
| //#define CV_8S 1 | |
| //#define CV_16U 2 | |
| //#define CV_16S 3 | |
| //#define CV_32S 4 | |
| //#define CV_32F 5 | |
| //#define CV_64F 6 | |
| //#define CV_16F 7 | |
| void printblob(InputArray blob_) { | |
| Mat blob = blob_.getMat(); |
Student: Zihao Mu
Mentor: Rostislav Vasilikhin
Link to accomplished work:
Student: Zihao Mu
Mentor: Vladimir Tyan
Link to accomplished work:
