jolibrain / livedetect Goto Github PK

View Code? Open in Web Editor NEW

50.0 50.0 9.0 9.5 MB

Live video client to DeepDetect

Home Page: https://www.deepdetect.com/

License: MIT License

Go 100.00%

livedetect's People

Contributors

Stargazers

Watchers

Forkers

margottd haijunma homerj233 jimmykarily alx tawawhite raceland-automacao-lda bycob iq-scm

livedetect's Issues

fix: display bounding box with different colors when predicting on multiple services

Now that livedetect can use prediction on multiple deepdetect service, livedetect should be able to display the bounding boxes (or masks) from these services.

Currently, only bounding boxes from the last service prediction is displayed.

Bounding boxes from each service should have a distinct color, a qualitative color palette could be use to differentiate these services: http://colorbrewer2.org/#type=qualitative&scheme=Set1&n=9

Can't access to the preview

Bonjour,

Config :

Rasperry B+ with Raspian Buster with desktop (release : 2019-07-10, kernel : 4.19)
DeepDetect and LiveDetect easily installed as explain in this tutorial : https://github.com/jolibrain/livedetect/wiki/Step-by-step-for-Raspberry-Pi-3

As written in the doc, DeepDetect is launched with :
docker run -d -p 8080:8080 -v $HOME/models:/opt/models jolibrain/deepdetect_ncnn_pi3

DeepDetect is running :

{
"status":{"code":200,"msg":"OK"},
"head":"method":"/info",
"version":"0.1",
"branch":"master",
"commit":"c1fdd1da53d4801bc7048f9f7f58699cf80c8c5d",
"services":[]}
}

As written in the doc, LiveDetect has been initiated with :

./livedetect-rpi3 \
    --port 8080 \
    --host 127.0.0.1 \
    --mllib ncnn \
    --width 300 --height 300 \
    --detection \
    --create --repository /opt/models/voc/ \
    --init "https://www.deepdetect.com/models/init/ncnn/squeezenet_ssd_voc_ncnn_300x300.tar.gz" \
    --confidence 0.3 \
    -v INFO \
    -P "0.0.0.0:8888" \
    --service voc \
    --nclasses 21

I start LiveDetect with :

./livedetect-rpi3 \
    --port 8080 \
    --host 127.0.0.1 \
    --mllib ncnn \
    --width 300 --height 300 \
    --detection \
    --init "https://www.deepdetect.com/models/init/ncnn/squeezenet_ssd_voc_ncnn_300x300.tar.gz" \
    --confidence 0.3 \
    -v INFO \
    -P "0.0.0.0:8888" \
    --service voc \
    --nclasses 21

LiveDetect starts like a charm :

[✔] [INFO] Starting capture
[✔] [INFO] Device: /dev/video0
[✔] [INFO] Starting web preview on 0.0.0.0:8888
================================================================================
[✖] [WARNING] Model size specified as parameters can't be use for capture with this camera.
[✖] [WARNING] Input image will be resized during DeepDetect processing.
[✔] [1] [INFO] Processing image 2019-07-28-12-01-39
[✔] [1] [INFO] Picture processed in 129.597491ms
================================================================================
[✔] [2] [INFO] Processing image 2019-07-28-12-01-39
[✔] [2] [INFO] Picture processed in 34.107204ms
================================================================================
[✔] [3] [INFO] Processing image 2019-07-28-12-01-39
[✔] [3] [INFO] Picture processed in 32.838772ms
================================================================================
[✔] [4] [INFO] Processing image 2019-07-28-12-01-39
[✔] [4] [INFO] Picture processed in 32.875542ms
================================================================================
[✔] [5] [INFO] Processing image 2019-07-28-12-01-39
[✔] [5] [INFO] Picture processed in 29.66092ms
================================================================================
[✔] [6] [INFO] Processing image 2019-07-28-12-01-39
[✔] [6] [INFO] Picture processed in 34.371006ms

Issue :
The doc says : "The preview is displayed on port 8888, that can be modified with flag". But I can't access in a browser neither with http://localhost:8888, 127.0.0.1:8888, 0.0.0.0:8888 as URL.

This line appears in the terminal :
2019/07/28 12:20:04 Stream: 192.168.1.25:53740 connected

But the browser still waiting during hours, whithout any information, neither json or video.

$HOME/models is root:root

Hi,
Thanks for this great project

There's just an issue in the instructions when using deepdetect with docker

docker run -d -p 8080:8080 -v $HOME/models:/opt/models jolibrain/deepdetect_cpu

if ~/models doesn't exist it gets created by docker as root:root and dede can't write to it (because it's executed as 1000:1000) which results in

[2019-05-16 12:51:15.318] [api] [info] Downloading init model https://www.deepdetect.com/models/init/ncnn/squeezenet_ssd_voc_ncnn_300x300.tar.gz
[2019-05-16 12:51:15.824] [api] [error] 172.17.0.1 "PUT /services/voc" 400 506
[2019-05-16 12:51:15.945] [api] [error] 172.17.0.1 "POST /predict" 400 0

Workarounds:

docker run -d -p 8080:8080 jolibrain/deepdetect_cpu
Precreate & chown ~/models:

mkdir ~/models
sudo chown 1000:1000 ~/models
docker run -d -p 8080:8080 -v $HOME/models:/opt/models jolibrain/deepdetect_cpu

The best would be to have deepdetect_cpu's entrypoint to chown by itself

Cheers,
Pierre

Installation fails due to `influxdb/client/v2`

Error is below:

 go get -u github.com/jolibrain/livedetect
package github.com/influxdata/influxdb/client/v2: cannot find package "github.com/influxdata/influxdb/client/v2" in any of:
        /usr/local/go/src/github.com/influxdata/influxdb/client/v2 (from $GOROOT)
        /home/pi/go/src/github.com/influxdata/influxdb/client/v2 (from $GOPATH)

Installation fails due to `influxdb/client/v2`

I have this error while installing influxdb

[root@mq gowork]# go get -u github.com/influxdata/influxdb/client/v2
package github.com/influxdata/influxdb/client/v2: cannot find package "github.com/influxdata/influxdb/client/v2" in any of:
/usr/lib/golang/src/github.com/influxdata/influxdb/client/v2 (from $GOROOT)
/opt/gowork/src/github.com/influxdata/influxdb/client/v2 (from $GOPATH)
[root@mq gowork]#

fix: save image on disk right after camera capture

Image is currently saved after deepdetect processing.

It should be saved right after the capture buffer is received from the camera.

Recommend Projects

React

A declarative, efficient, and flexible JavaScript library for building user interfaces.
Vue.js

🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
Typescript

TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
TensorFlow

An Open Source Machine Learning Framework for Everyone
Django

The Web framework for perfectionists with deadlines.
Laravel

A PHP framework for web artisans
D3

Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

javascript

JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
web

Some thing interesting about web. New door for the world.
server

A server is a program made to process requests and deliver data to clients.
Machine learning

Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Visualization

Some thing interesting about visualization, use data art
Game

Some thing interesting about game, make everyone happy.

Recommend Org

Facebook

We are working to build community through open source technology. NB: members must have two-factor auth.
Microsoft

Open source projects and samples from Microsoft.
Google

Google ❤️ Open Source for everyone.
Alibaba

Alibaba Open Source for everyone
D3

Data-Driven Documents codes.
Tencent

China tencent open source team.