VisionAl is a open-source ready-to-use Python Library for various Computer Vision Scenarios You can install it through ‘pip install visional You can browse different scenarios supported at https://docs.visionify.ai/scenarios/

VisionAl library is focused on common workplace and employee health & safety scenarios You can browse various scenarios we support at https://docs.visionify.ai/scenarios/. At a high-level these include the employee health and safety hazard & fire warnings, equipment monitoring, vehicle monitoring people & productivity monitoring, auspicious activity monitoring and common company compliance policies.

No! You do not need any new comeara or hardware to run this system VisionAl works with your existing security camera infrastructure. We support RTSP, RTMP, HLS and other common video platforms Current safety surveillance systems are just record and playback – we can bring a lot of operational and safety insights from the current camera systems That said, some of VisionAl scenarios are based on IR and ToF Cameras And still others are loT sensors. Based on the scenario you want to run – we can customize the camera and loT censors configuration

We are a Open-Source first company All our Software is publicly available and easy to download online You can get the library source at https://github.com/visionify/visionai. Our software is licensed under GPLv3 License This means that any applications you link this software to have to be open-sourced under GPLv3 as well.

We recommend testing this on a beefy machine with a NVIDIA graphics card. To quickly test out a scenario you can follow these commands:

————
pip install visional

visional web

————

And then browse the different scenarios, create pipelines for our organization. We also provide a Azure Managed App which has all the dependencies pre-installed. You can get it here

We work with our clients to quickly create customized models for their use-cases. This is available to be purchased through Azure marketplace as a Consulting Service-we recommend this option for a quicker transaction. You can find more details here.

We take our customers data privacy very seriously. All our current models available in the community edition are based off of open-source datasets We have several customer specific models trained on private data, but those datasets are maintained on their own storage accounts.
They are not used for training any publicly available models.
This way we are able to serve open-source community as well as commercial use-cases.

Our system is isolated on its own and all it needs is a set of camera output. We are focused on building more use-cases to cover common safety and compliance scenarios. As such, we would work with you to build a framework where these use cases are isolated and easy to. integrate into your organization.

Our license terms are flexible to provide your in-house ML team a starting point where they can build their new use-cases. We also provide a customer success team that can with you to understand your requirements and guide you in coming up with the right solutions for the problems you are working on.