Getting Started

Signup and Login:

1. New User - Register at Qeexo.com for a free account

2. Existing Users: Login to automl.qeexo.com

Installation

1. Qeexo AutoML Desktop Client: From Resources, download and install the Qeexo AutoML Desktop client

2. Devices and Drivers: Install your device and drivers

Project Creation

1. From the Project Creation window

   1. Give your project a Project Name

   2. Select a 'Classification Type'

   3. Select a supported hardware device

   4. Click Create

Data Collection

1. From the Data Collection window

   1. Provide your environment an Environment Name

   2. Click Edit to select and Configure Sensors

   3. Collection Name

   4. Collection Duration (seconds)

   5. Click Record

2. From the Data Recording Window:

   1. Click Start to collect data

   2. Repeat steps, 6.a.iii to 6.b.i to collect different data collection

Start Training

1. From the Data Page--> Training data, select the data collection collected or uploaded

   1. Group Labels - Optional

   2. Select Sensors - Automatic or Manual

   3. Select Inference Settings - Automatic or Manual

   4. Select Algorithms

   5. Validate Model Settings( Automatically determined)

   6. Start Training

2. Data Visualization

   1. From the Data Page, Click on Visualization

   2. Visualize data based on sensor types and plot types

3. Data Segmentation

   1. From the Data Page, click on the ellipse (three dots) to select Segment option

   2. Once open, use the drag bars and your mouse pointer to locate and zoom a region of interest

   3. Next click CREATE NEW SEGMENT to enter desired segment label and select segment color, click create

   4. Locate an event on the graph and click its start point and end point to create segment(s)

4. Check Training Model’s Progress & Performance

   1. Click on Models tab to check model's details and build progress

   2. When the build status is ‘READY’ , click on Training Results to see the model’s performance

Live Testing

1. From the Models tab, select the Training Results to evaluate the models we built

2. Select PUSH TO HARDWARE to push the best-performing model to the device for Live Testing

3. Wait for Flashing to be successful

4. Do Live Testing