Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Efficient monocular visual odometry for ground vehicles on ARM processors
| Date | Stars |
|---|---|
| 2026-07-24 | 256 |
| 2026-07-25 | 256 |
| 2026-07-28 | 256 |
| 2026-07-30 | 256 |
| 2026-08-06 | 256 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
15.0
growth rate 0.00%/day
# ARM-VO ARM-VO is a monocular visual odometry algorithm designed for on-road vehicles. It is highly optimized for ARM CPUs as it uses NEON C intrinsics and multi-threading to accelerate keypoint detection and tracking. ## Results on KITTI dataset | Sequence 05 | Sequence 07 | Sequence 10 | |:---:|:---:|:---:| | <img src="docs/assets/Sequence5.png" width="100%"> | <img src="docs/assets/Sequence7.png" width="100%"> | <img src="docs/assets/Sequence10.png" width="100%"> | ## What's new in v2? - Results are deterministic - Scale estimation is more accurate (but slower) - Camera pitch angle is no longer required (providing camera height is enough) - RGB and BGR inputs are supported - Distorted images are supported - Keypoint tracking is faster by re-using KLT pyramids - Motion estimation is more robust in dynamic environments - The API and the implementation are much cleaner - Enabled compilation on x86 machines to simplify development - Removed ROS node examples (will be back in future) ## Dependencies - C++17 (or above) - CMake >= 3.20 and build essentials ```bash sudo apt install build-essential git cmake pkg-config ``` - OpenCV ```bash git clone --branch 4.10.0 --depth 1 https://github.com/opencv/opencv.git cd opencv mkdir build && cd build cmake -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=/usr/local -DBUILD_TESTS=OFF -DBUILD_PERF_TESTS=OFF -DBUILD_DOCS=OFF -DBUILD_EXAMPLES=OFF -DENABLE_NEON=ON -DBUILD_opencv_python2=OFF -DBUILD_opencv_python3=OFF .. make -j$(nproc) sudo make install sudo ldconfig ``` - ncnn ```bash git clone --recursive --depth 1 --branch 20241226 https://github.com/Tencent/ncnn.git cd ncnn mkdir build && cd build cmake -DCMAKE_BUILD_TYPE=Release -DCMAKE_CXX_STANDARD=17 -DCMAKE_INSTALL_PREFIX=/usr/local -DNCNN_BUILD_TESTS=OFF -DNCNN_BUILD_EXAMPLES=OFF -DNCNN_BUILD_BENCHMARK=OFF -DNCNN_THREADS=ON -DNCNN_OPENMP=OFF -DNCNN_VULKAN=ON -DNCNN_ENABLE_LTO=ON .. make -j$(nproc) sudo make install sudo ldconfig ``` - Catch2 (only if you want to build tests as well) ```bash git clone --branch v2.13.10 --depth 1 https://github.com/catchorg/Catch2.git cd Catch2 mkdir build && cd build cmake -DCMAKE_BUILD_TYPE=Release -DCMAKE_CXX_STANDARD=17 -DCMAKE_INSTALL_PREFIX=/usr/local -DCATCH_BUILD_STATIC_LIBRARY=ON -DCATCH_BUILD_TESTING=OFF -DCATCH_INSTALL_DOCS=OFF -DCATCH_INSTALL_HELPERS=ON .. make -j$(nproc) sudo make install sudo ldconfig ``` ## How to build? ```bash git clone https://github.com/zanazakaryaie/ARM-VO.git cd ARM-VO mkdir build && cd build cmake -DCMAKE_BUILD_TYPE=Release .. make -j$(nproc) sudo make install sudo ldconfig ``` ### Build Options | Option | Default | Description | |---|---:|---| | `BUILD_TESTS` | `OFF` | Build unit tests | ## Run on KITTI dataset Download the odometry dataset from [here](https://s3.eu-central-1.amazonaws.com/avg-kitti/data_odometry_color.zip). Open a terminal, navigate to build/cli folder and run: ```bash ./run_armvo --image_folder=path/to/downloaded/images/folder --config=path/to/config.yaml ``` To compare ARM-VO's accuracy with ground-truth poses, first download the ground-truth data from [here](https://s3.eu-central-1.amazonaws.com/avg-kitti/data_odometry_poses.zip). Then navigate to build/cli folder and run: ```bash ./run_armvo --image_folder=path/to/downloaded/images/folder --config=path/to/config.yaml --gt_poses=path/to/ground-truth/poses/foo.txt ``` ## How to use ARM-VO in your project? If your project uses CMake, you can find the installed ARM-VO package and link against the core visual odometry library: ```cmake find_package(armvo REQUIRED CONFIG) target_link_libraries(my_app PRIVATE armvo::ArmVO) ``` The package also exports `armvo::ArmVOtools` for helper utilities. Link it if your application needs the tools API: ```cmake find_package(armvo REQUIRED CONFIG) target_link_libraries(my_app PRIVATE armvo::ArmVO armvo::ArmVOtools) ``` ## Limitations - ARM-VO recovers the scale if 1) the camera height i
Excerpt of 5,991 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:cee95919b473b44f, topic:opencv
matched fp:cee95919b473b44f, topic:ros