Build interactive AR apps with our real-time 3D hand tracking & gesture recognition technology.
Hand Detection and Tracking
Our neural network automatically and accurately detects and captures hand movements in 3D. Its algorithm is trained on a dataset of images with different lighting, hand positions, male and female hands. It makes our hand tracking software run smoothly in real-world applications and almost on any mobile device.
We build a hand skeleton model of 11 points on fingers, fingertips, palm, and wrist. It lets us estimate and accurately predict the position of a hand and fingers, even if they are partially overlapped. Users can naturally move hands without breaking an AR experience.
With gesture recognition, your users can trigger AR experience or control interfaces with gestures. You can build remotely controller photo booths and magic mirrors that are interactive, fun, and absolutely safe.
- Automatic real-time hand detection
- Distance: 2 meters
- Left and right hands recognized
- Stable output with different hand angles and movements
- Adapted to low lighting, outdoor or in-store use cases
- Minimal power consumption on mobile devices
- Compatible with Face AR and background removal
Take mobile AR interactions to the next level with our hand tracking software easily added to your iOS and Android apps.
Bring fun to video chats letting users virtually touch each other, throw objects, or trigger animation with gestures.
Boost sales by letting customers try rings on in AR.
Build touchless interfaces that are natural, hygienic, and memorable.
Create AR characters and control them with your hands, just like real puppets.
Technical Features & System Requirements
Why Banuba’s Technology
- 100% proprietary technology and published patents
- Guaranteed data privacy and compliance with GDPR due to offline work
- Technology development and optimization for client’s hardware
- Own R&D center specifically focused on Face AR
- 7-year team expertise in AR content development and design
- Trusted by global brands, enterprise partners like Gucci, Sberbank, Yandex
- Fast support and team member consultancy
- Hands-on experience in face tracking app development and release
Detecting a human hand in the picture and determining its exact location. Hand tracking can be used in eCommerce, video conferencing, entertainment, and other industries.
Hand tracking works similarly in both augmented and virtual reality. The specially trained neural networks label each pixel in the image and determine whether it belongs to a hand or not.
You can either build a hand tracking system from scratch or use a premade SDK. If you are good at computer vision or have a team on hand, you can DIY. However, this would likely take a long time, so SDKs are more efficient in this regard.