SSD Mobilenet model for sale
Introducing the new SSD Mobilenet model – the latest in human detection technology for manufacturing environments!
Our model is built on 3000 images and approximately 5000 annotations. It is designed to be used both on Jetson and CPU platforms. The model has an impressive resolution of 1920x1080 and 4K images. Allowing it to detect people wearing different helmets and clothing colors. It can detect humans from a variety of angles, including topview, closeview, far away and also while crouching or if only the upper body is visible.
Story of the model
The model is trained to detect humans in danger areas. Also to detect forklifts near production lines. With a high accuracy and low classification and regression loss, this model is a powerful tool for any industrial setting. It is trained for 2000 epochs, taking about 5 days. We used dual GPU, and is tested in production environments. Currently it is used in industrial enviroment to detect humans in danger area.
High resolution model
The 512x512 resolution of the model is even better than regular SSD 300x300. It has low detection time, making it perfect to use on Jetson. Our model comes with all the classes you need – human, closebody, topbody, crouch, upperbody and forklift.
Try the SSD Mobilnet model today and experience the power of human detection in manufacturing environments!
Here you can download our free trial SSD Mobilenet model, which is trained only 100 epochs, has less input images and annotations. Download free SSD Mobilenet for Industrial Enviroment here.
If that model is suitable for you, contact us and let’s make a reasonable price for a model which has following qualities:
- 3000 input images
- 5000 annotations
- 512 x 512 pixels SSD Mobilenet V2
- Trained 2000 epochs
- Multiple classes: human, topbody, upperbody, crouching, closebody, forklift
- low classification loss
- low regression loss
- high accuracy, high precision
- Can be used on Jetson
We also provide services to even improve the model for your enivroment, also we can give you the model in other format.
Contact us to get model, where we have trained more than 2000 epochs and which includes more input images and annotations. raul.orav@visioline.ee or +372 504 9966
When creating the model, we made pictures of people with different colored clothing, with and without helmets, with work clothes and regular clothing, in different poses and in different distances and viewpoints. On the identification of the forklift, we trained with several different forklifts, positions and distances.
You want us to finish the whole project, take a look here: https://www.visioline.ee/people-and-equipment-detection/
SSD Mobilenet precision
SSD Mobilenet loss
Why to buy model?
When it comes to harnessing the power of machine learning in industrial environments, the most efficient approach is to buy a pre—trained model. Collecting a dataset and annotating it, running multiple tests and incrementally improving the model can be a time—consuming and complex process.
Investing in a pre—trained model can save you time and money, allowing you to focus on the aspects of your business that need your attention most. With a pre—trained model, you can be sure that you are using the most up—to—date and accurate machine learning technology and that you are getting the best possible results.
Buying a pre—trained model is the smarter option when it comes to leveraging the power of machine learning in industrial settings.
P.S. we have built different machine vision solutions — ask and we will provide you a reasonable price ;)