DeepStack UI — Object detection with zero code

This article is a step-by-step tutorial on how to use DeepStack-UI with zero-code for object detection.

In this tutorial, we are going to guide you on how to perform object detection in images from your web browser without writing a single code, using DeepStack and DeepStack UI.

What is DeepStack ?

DeepStack is an open-source AI API server that empowers developers, IoT experts, research teams and individuals in small and large companies around the world to easily deploy AI systems both on premise and in the cloud. DeepStack is device and programming language agnostic. You can run it on Windows, Mac OS, Linux, Raspberry PI ( + all ARM devices)and NVIDIA Jetson devices with CPU and GPU acceleration. The following are features in-built into DeepStack

• Face detection, matching and recognition APIs
• Common objects detection and recognition APIs
• Scene recognition APIs
• Custom objects detection APIs
• API security

DeepStack Documentation! | DeepStack v1.2.1 documentation

What is DeepStack-UI?

Deepstack-UI is a interactive web application built by Robin Cole . It allows anyone to run any image through DeepStack’s object detection API in a user-friendly web interface and customize your input/output without writing any code. You can select a specific section of the image on which you want objects will detected from, filter the object detected by their confidence and set the target object(s) .

GitHub - robmarkcole/HASS-Deepstack-object: Home Assistant custom component for using Deepstack…

How to install Deepstack-UI ?

To run Deepstack-UI for object detection, the 3 major steps below must be followed.

Step 1 — Install Docker

Install the version of Docker on your machine that corresponds to your operating system.

Install Docker Desktop on Windows

Install Docker Desktop on Mac

Install Docker Engine on Ubuntu

Step 2— Deepstack Installation on Docker

To install and run Deepstack, run the command below in PowerShell/CMD (Windows) or Terminal (Linux or MacOS).

docker run -e VISION-DETECTION=True -p 80:5000 deepquestai/deepstack:latest

Step3 — DeepStack-UI Installation

Deepstack-UI can be installed through three different ways

Option 1:

• clone DeepStack UI repository here
• cd into the repository
• Build the docker image locally with the command

docker build -t deepstack-ui . -f Dockerfile

• Run the DeepStack-UI Docker app

 docker run -p 8501:8501 -e DEEPSTACK_IP=localhost deepstack-ui

Option 2 :

Run the existing DeepStack-UI Docker application from Docker Hub

docker run -p 8501:8501 -e DEEPSTACK_IP=localhost robmarkcole/deepstack-ui:latest

Option 3:

• Clone the deepstack_ui repository
• Open Powershell, CMD or terminal in the repository folder and run pip install -r requirements.txt
• For Windows users, go to this folder on your laptop C:\Users<your-laptop-username>.streamlit/ and delete the file config.toml
• cd into the app folder of the repository
• Run the command below to start DeepStack-UO

streamlit run deepstack-ui.py

Visithttp://localhost:8501 on your web browser and you will see the interface below.

DeepStack-UI interface

Yeah, Congratulations ! Deepstack-UI is successfully running on your device. You can now select any image from your machine and DeepStack-UI will process it for object detection. Let’s use the sample image below.

sample Image

DeepStack-UI interface

After inputting an image into the web app, It profides a page similar to the one below

The landing page of DeepStack-UI consist of different sections that will ease your navigation on the application. It is composed of two sections; the parameter section constructed to manipulate the image as you desire and the main page which displayed the output results.

The parameter section is composed of 3 parts:

1. Confidence: this is used to set the threshold of the each object prediction percentage (confidence).
2. ROI : the Region of Interest is used to select a specific part of the image on which the detection will run.
3. Object Selection : This allows the user to choose the type of object to extract from the image.

The main page outputs four elements which are:

1. Processed Image : This image is composed of all object detected, bordered with their respective label
2. All discovered objects : A list of all object detected.
3. Filtered object count : A list of the objects and their number of occurrence.
4. All filtered objects : A list of dictionaries where each dictionary consist of each object’s confidence value, name and coordinates details.

Deepstack-ui application demonstration

This gif illustrates how to use all the attributes of Deepstack-ui to control an image.

For more details on using DeepStack-UI , visit the GitHub repository for the project via the link below.

GitHub - robmarkcole/deepstack-ui: UI for working with Deepstack

Do you enjoy using DeepStack-UI? If yes, share this article and also give the project a star on GitHub.

To read more on DeepStack AI Server, visit out blog via the link below

DeepQuestAI