You can train object detector model.
You can upload dataset of the formats given below: images – jpg, jpeg, png video – mp4
A minimum of 320×240 and a maximum of 2048×1536.
To create a model, Minimum 120 images are needed.
Less than 1GB of video file size is recommended.
The duration should be at least 2 minutes.
You can add a maximum of 10 labels.
There are two purposes as follows: – 1. You can continue to annotate the pending images. 2. Editing and training annotated images are possible.
You can upload dataset of the formats given below: images – jpg, jpeg, png video – mp4
The duration should be at least 2 minutes.
A minimum of 320×240 and a maximum of 2048×1536.
To create a model, Minimum 120 images are needed.
Less than 1GB of video file size is recommended.
.mp4 files are supported.
One model at a time is allowed.
A minimum of 320×240 and a maximum of 2048×1536.
Less than 1GB of video file size is recommended.
jpg, jpeg, png
4GB – 1 model
6GB – 2 models
8GB – 3 models
12GB – 4 models
Yes, You can download the images
The application is free and offline and does not require an internet connection to run (internet connection is required only during installation and one-time setup).
All datasets and models are built on your machine locally, so you are in control of your data. Your data is never uploaded to the cloud.
Yes, by clicking on Projects –> click on download icon for the model file you wish to download.
The exported model file can be found in ‘Downloads’.
Yes.
It supports Japanese, Hindi, Marathi, English, and Gujrati
You can train 1 model at a time.
Chrome, Firefox.
Yes, when the user starts the software for the 1st time.