{"id":21896754,"url":"https://github.com/srohit0/trafficvision","last_synced_at":"2026-03-10T06:04:54.389Z","repository":{"id":264521368,"uuid":"164944430","full_name":"srohit0/trafficVision","owner":"srohit0","description":"MIVisionX toolkit is a comprehensive computer vision and machine intelligence libraries, utilities and applications bundled into a single toolkit.","archived":false,"fork":false,"pushed_at":"2019-06-26T22:54:55.000Z","size":15892,"stargazers_count":67,"open_issues_count":2,"forks_count":22,"subscribers_count":8,"default_branch":"master","last_synced_at":"2025-04-15T17:56:23.206Z","etag":null,"topics":["amd-gpu","amd-modules","amd-opencl","amd-openvx","artificial-intelligence","artificial-neural-networks","convolutional-neural-networks","machine-intelligence","machine-learning","mivision","mivisionx","object-detection","opencl","openvx","openvx-nn-extension","rocm","tiny-yolo","tiny-yolo-network","yolov2"],"latest_commit_sha":null,"homepage":"https://gpuopen-professionalcompute-libraries.github.io/MIVisionX/","language":"C++","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/srohit0.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null}},"created_at":"2019-01-09T21:57:22.000Z","updated_at":"2024-10-22T16:35:33.000Z","dependencies_parsed_at":null,"dependency_job_id":"93d6b5e9-6877-47cf-8ae0-7427fee670b0","html_url":"https://github.com/srohit0/trafficVision","commit_stats":null,"previous_names":["srohit0/trafficvision"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/srohit0/trafficVision","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/srohit0%2FtrafficVision","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/srohit0%2FtrafficVision/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/srohit0%2FtrafficVision/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/srohit0%2FtrafficVision/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/srohit0","download_url":"https://codeload.github.com/srohit0/trafficVision/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/srohit0%2FtrafficVision/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":30326878,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-03-10T05:25:20.737Z","status":"ssl_error","status_checked_at":"2026-03-10T05:25:17.430Z","response_time":106,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["amd-gpu","amd-modules","amd-opencl","amd-openvx","artificial-intelligence","artificial-neural-networks","convolutional-neural-networks","machine-intelligence","machine-learning","mivision","mivisionx","object-detection","opencl","openvx","openvx-nn-extension","rocm","tiny-yolo","tiny-yolo-network","yolov2"],"created_at":"2024-11-28T14:12:17.384Z","updated_at":"2026-03-10T06:04:54.381Z","avatar_url":"https://github.com/srohit0.png","language":"C++","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Traffic Vision\nThis app detects cars/buses in a live traffic at a phenomenal **50 frames/sec with HD resolution (1920x1080)** using deep learning network [Yolo-V2](https://pjreddie.com/darknet/yolov2). The model used in the app is optimized for inferencing performnce on AMD-GPUs using [MIVisionX toolkit](https://gpuopen-professionalcompute-libraries.github.io/MIVisionX/).\n\n[![Traffic Vision Animation](media/traffic_viosion.gif)](https://youtu.be/YASOovwds_A)\n\n## Features\n1. Vehicle detection with bounding box\n1. Vehicle direction ((upward, downward) detection \n1. Vehicle speed estimation\n1. Vehicle type: bus/car.\n\n## How to Run\n\n### Use Model\n\u003cimg src=\"media/speed_detection_user_interface.jpg\" width=600\u003e\n\n### Demo\n\nApp starts the demo, if no other option is provided. Demo uses a video stored in the [media/](./media) dir.\n```\n% ./main.py\n('Loaded', 'yoloOpenVX')\nOK: loaded 22 kernels from libvx_nn.so\nOK: OpenVX using GPU device#0 (gfx900) [OpenCL 1.2 ] [SvmCaps 0 1]\nOK: annCreateInference: successful\nProcessed a total of  102 frames\nOK: OpenCL buffer usage: 87771380, 46/46\n%\n```\nHere is the [link to YouTube video](https://youtu.be/YASOovwds_A) detecting cars, bounding boxes, car speed, and confidence scores.\n### Other Examples\n\n**recorded video**\n\u003e 1. ./main.py --video \u003cpath-to-video\u003e/vid.mp4\n  \n**traffic cam ip** \n\u003e 2. ./main.py --cam_ip 'http://166.149.104.112:8082/snap.jpg'\n\n## Installation\n\n### Prerequisites \n\n1. GPU: Radeon Instinct or Vega Family of Products with [ROCm](https://rocm.github.io/ROCmInstall.html) and OpenCL development kit\n1. [Install AMD's MIVisionX toolkit](https://gpuopen-professionalcompute-libraries.github.io/MIVisionX/) : AMD's MIVisionX toolkit is a comprehensive computer vision and machine intelligence libraries, utilities\n1. [CMake](http://cmake.org/download/), [Caffe](http://caffe.berkeleyvision.org/installation.html)\n1. [Google's Protobuf](https://github.com/google/protobuf)\n\n### Steps\n\n```\n% git clone https://github.com/srohit0/trafficVision\n```\n\n\n**_1. Model Conversion_**\n\nThis steps downloads yolov2-tiny for voc dataset and converts to MIVision's openVX model. \n```\n% cd trafficVision/model\n% bash ./prepareModel.sh\n```\nMore details on the pre-requisite (like [caffe](http://caffe.berkeleyvision.org/installation.html)) of the model conversion in the [models/](./models) dir.\n\n**_2. MIVision Model Compilation_**\n\n```\n% cd trafficVision\n% make\n```\n\n**_3. Test App_**\n\n```\n% cd trafficVision\n% make test\n```\nIt'll display detection all videos in media/ dir.\n\n## Design\nThis section is a guide for developers, who would like to port vision and object detections models to AMD's Radeon GPUs from other frameworks including [tensorflow](https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/detection_model_zoo.md), [caffe](http://caffe.berkeleyvision.org/model_zoo.html) or [pytorch](https://pytorch.org/).\n\n### High Level Design\n\u003cimg src=\"media/speed_detection_top_level_arch.jpg\" width=\"600\"\u003e\n\n### Lower Level Modules\nThese lower level modules can be found as python modules (files) or packages (directories) in this repository.\n\u003cimg src=\"media/speed_detection_modules.jpg\" width=\"600\"\u003e\n\n## Development\n\n### Model Conversion\nFollow model conversion process similar to the one described below.\n\u003cimg src=\"media/speed_detection_model_conversion.jpg\" width=680\u003e\n\n\n### Infrastructure\nMake sure you've infrastructure pre-requisites installed before you start porting neural network model for inferencing.\n\u003cimg src=\"media/speed_detection_infrastructure.jpg\" width=480\u003e\n\n## Developed and Tested on\n1. Hardware\n    1. AMD Ryzen Threadripper 1900X 8-Core Processor\n    1. Accelerator = Radeon Instinct™ MI25 Accelerator \n1. Software\n    1. Ubuntu 16.04 LTS OS\n    1. Python 2.7\n    1. MIVisionX 1.7.0\n    1. AMD OpenVX 0.9.9\n    1. GCC 5.4\n\n## Credit\n* MIVisionX Team\n\n## References\n1. [yoloV2 paper](https://arxiv.org/pdf/1612.08242.pdf)\n1. [Tiny Yolo aka Darknet reference network](https://pjreddie.com/darknet/imagenet/#reference)\n1. [MiVisionX Setup](https://github.com/kiritigowda/MIVisionX-setup)\n1. [AMD OpenVX](https://gpuopen.com/compute-product/amd-openvx/)\n1. [Optimization with OpenVX Graphs](http://openaccess.thecvf.com/content_cvpr_workshops_2014/W17/papers/Rainey_Addressing_System-Level_Optimization_2014_CVPR_paper.pdf)\n1. [Measuring Traffic Speed With Deep Learning Object Detection](https://medium.com/datadriveninvestor/measuring-traffic-speed-with-deep-learning-object-detection-efc0bb9a3c57)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsrohit0%2Ftrafficvision","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fsrohit0%2Ftrafficvision","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsrohit0%2Ftrafficvision/lists"}