README.md

👀 yolo - You Only Look Once service

Provides entity count information as micro-service; updates periodically (default 0 seconds). This container may be run locally using Docker, pushed to a Docker registry, and published to any Open Horizon exchange.

Status

CPU only

GPU accelerated

Service discovery

• org - github@dcmartin.com
• url - com.github.dcmartin.open-horizon.yolo
• version - 0.1.2

Service variables

• YOLO_CONFIG - configuration of YOLO; tiny-v2, tiny-v3,v2, or v3
• YOLO_ENTITY - entity to count; defaults to all
• YOLO_PERIOD - seconds between updates; defaults to 0
• YOLO_THRESHOLD - minimum probability; default 0.25; range 0.0 to 1.0
• LOG_LEVEL - specify level of logging; default info; options include (debug and none)
• DEBUG - turn on debugging output; true or false; default false

GPU-enabled versions

There are two version of this service which are enabled with nVidia GPU capabilities, drastically reducing the time required to perform a prediction.

• yolo-cuda - For AMD/Intel 64-bit systems with nVidia GPU supporting CUDA version 10
• yolo-tegra - For nVidia Jetson Nano and other tegra devices with JetPack v3.2

For both services the default container run-time for Docker must be the nvidia-container-runtime. The /etc/docker/daemon.json file must contain the following:

{
  "default-runtime": "nvidia",
  "runtimes": {
    "nvidia": {
      "path": "/usr/bin/nvidia-container-runtime",
      "runtimeArgs": []
    }
  }
}

How To Use

Copy this repository, change to the yolo directory, then use the make command; see below:

% mkdir ~/gitdir
% cd ~/gitdir
% git clone http://github.com/dcmartin/open-horizon
% cd open-horizon/yolo
% make
...
{
  "yolo": null,
  "date": 1554316177,
  "hzn": {
    "agreementid": "",
    "arch": "",
    "cpus": 0,
    "device_id": "",
    "exchange_url": "",
    "host_ips": [
      ""
    ],
    "organization": "",
    "ram": 0,
    "pattern": null
  },
  "config": {
    "log_level": "info",
    "debug": false,
    "date": 1554316177,
    "period": 0,
    "entity": "all",
    "scale": "none",
    "config": "tiny",
    "device": "/dev/video0",
    "threshold": 0.25,
    "services": null,
    "names": [
      "person", "bicycle", "car", "motorbike", "aeroplane", "bus", "train", "truck", "boat", "traffic light", "fire hydrant", "stop sign", "parking meter", "bench", "bird", "cat", "dog", "horse", "sheep", "cow", "elephant", "bear", "zebra", "giraffe", "backpack", "umbrella", "handbag", "tie", "suitcase", "frisbee", "skis", "snowboard", "sports ball", "kite", "baseball bat", "baseball glove", "skateboard", "surfboard", "tennis racket", "bottle", "wine glass", "cup", "fork", "knife", "spoon", "bowl", "banana", "apple", "sandwich", "orange", "broccoli", "carrot", "hot dog", "pizza", "donut", "cake", "chair", "sofa", "pottedplant", "bed", "diningtable", "toilet", "tvmonitor", "laptop", "mouse", "remote", "keyboard", "cell phone", "microwave", "oven", "toaster", "sink", "refrigerator", "book", "clock", "vase", "scissors", "teddy bear", "hair drier", "toothbrush"
    ]
  },
  "service": {
    "label": "yolo",
    "version": "0.0.8"
  }
}

The yolo payload will be incomplete until the service completes; subsequent make check will return complete; see below:

{
  "yolo": {
    "mock": "eagle",
    "info": {
      "type": "JPEG",
      "size": "773x512",
      "bps": "8-bit",
      "color": "sRGB"
    },
    "time": 0.861101,
    "count": 1,
    "detected": [
      {
        "entity": "bird",
        "count": 1
      }
    ],
    "image": "<redacted>",
    "date": 1554316261
  },
  "date": 1554316177,
  "hzn": {
    "agreementid": "",
    "arch": "",
    "cpus": 0,
    "device_id": "",
    "exchange_url": "",
    "host_ips": [
      ""
    ],
    "organization": "",
    "ram": 0,
    "pattern": null
  },
  "config": {
    "log_level": "info",
    "debug": false,
    "date": 1554316177,
    "period": 0,
    "entity": "all",
    "scale": "none",
    "config": "tiny",
    "device": "/dev/video0",
    "threshold": 0.25,
    "services": null,
    "names": [
      "person", "bicycle", "car", "motorbike", "aeroplane", "bus", "train", "truck", "boat", "traffic light", "fire hydrant", "stop sign", "parking meter", "bench", "bird", "cat", "dog", "horse", "sheep", "cow", "elephant", "bear", "zebra", "giraffe", "backpack", "umbrella", "handbag", "tie", "suitcase", "frisbee", "skis", "snowboard", "sports ball", "kite", "baseball bat", "baseball glove", "skateboard", "surfboard", "tennis racket", "bottle", "wine glass", "cup", "fork", "knife", "spoon", "bowl", "banana", "apple", "sandwich", "orange", "broccoli", "carrot", "hot dog", "pizza", "donut", "cake", "chair", "sofa", "pottedplant", "bed", "diningtable", "toilet", "tvmonitor", "laptop", "mouse", "remote", "keyboard", "cell phone", "microwave", "oven", "toaster", "sink", "refrigerator", "book", "clock", "vase", "scissors", "teddy bear", "hair drier", "toothbrush"
    ]
  },
  "service": {
    "label": "yolo",
    "version": "0.0.8"
  }
}

Example

RaspberryPi Camera Module

The Pi’s camera has a discrete set of input modes. On the V2 camera these are as follows:

Number	Resolution	Aspect Ratio	Framerates	Video	Image	FoV	Binning
1	1920x1080	16:9	0.1-30fps	x			Partial
2	3280x2464	4:3	0.1-15fps	x	x	Full	None
3	3280x2464	4:3	0.1-15fps	x	x	Full	None
4	1640x1232	4:3	0.1-40fps	x			Full
5	1640x922	16:9	0.1-40fps	x			Full
6	1280x720	16:9	40-90fps	x			Partial
7	640x480	4:3	40-90fps	x			Partial

Field-of-View

Changelog & Releases

Releases are based on Semantic Versioning, and use the format of MAJOR.MINOR.PATCH. In a nutshell, the version will be incremented based on the following:

• MAJOR: Incompatible or major changes.
• MINOR: Backwards-compatible new features and enhancements.
• PATCH: Backwards-compatible bugfixes and package updates.

Authors & contributors

David C Martin (github@dcmartin.com)