Be at the forefront of AI solutions
which see in 3D
What is 3D Movement Recognition, and why is it needed?
3D Movement Recognition (also known as 3D Movement Analysis) is a method for evaluating human and machine movements in three-dimensional space. It is used to detect and track movement patterns and evaluate the efficiency of particular movements. 3D movement analysis can be used in a wide range of applications to optimize processes, improve performances, prevent injuries, and detect dangerous situations in different environments.
3D Point Cloud Data
To visualize 3D data, we often use point cloud data where the original geometric information is preserved in 3D space.We use Deep Learning, 3D Machine Learning and 3D Computer Vision to perform complex tasks such as recognition, detection and retrieval from the 3D point cloud data. With very good results. With each of our projects, we make sure to understand your needs in order to select the best techniques for your specific scenario.
3D Analysis Scenarios
Explore how you can leverage the potential of 3D Movement Recognition for your business, no matter if you want to detect dangerous situations, analyze clinical activities, or enable human-machine cooperation.
Detect dangerous situations
Detect dangerous situations such as patient or elderly person falling, recognize epileptic seizures, analyze illness-related motions.
Analyze clinical activities
Analyze and support clinical activities, e.g. analyze movements in the operating room and get specific information.
AI personal trainer or digital physiotherapist
Recognize physical exercises and movements, analyze their correctness and get personalized insights. Measure more features using 3D data than a single RGB camera.
Metaverse, gaming and VR
Use Microsoft Azure Kinect 3D cameras for markerless motion capture. Analyze the real objects and present them together with movements in form of 3D models in multiverse scenarios.
Industrial human-machine appliances
Precisely detect the movements of people in different environments where human and machines work together. Optimize the workflows and avoid accidents and errors.
Digital twin of environments in 3D
Use the capabilities offered by special 3D cameras to create a digital twin of your physical environment. Create a 3D digital twin to gain accurate insights and see what is happening in your facility. Analyze operations and activities, identify bottlenecks, and take action to improve operational efficiency in your business or clinic. Examples of environments include:
- Operating room or hospital
- Warehouse
- Construction sites
- Store or shopping mall
We are working with global hardware leaders
To ensure the success of our customers and solutions, we have established partnerships with leading technology providers such as Intel, Microsoft and Nvidia.
Key Features
Skeleton Estimation
Accurately estimate human pose through key points, based on video streams from multiple 3D and 2D cameras. Detect the position and movements of the body parts.
Measurements and Angles
Automatic detection of poses – e.g. whether angles between body parts are within the correct range, distances are correct or whether certain key points do not exceed the allowed range – such as whether the knees do not go over the toes.
Human Activity Recognition
Use the latest AI techniques to recognize and analyze specific activities, such as physical exercise, purchasing behavior, health-related movements like falls or epileptic seizures, or activities at assembly line or in a warehouse.
3D cameras and Edge AI
We implement the projects using best-in class 3D cameras (Azure Kinects) and newest Edge AI hardware from Nvidia and Intel. We have developed efficient methods for multi-camera stream processing which allow for effective data collection and processing directly on the edge devices.
Visualization
Play-back the recognized activities and human poses, visualize the results in 2D and 3D space or use the captured motions to mimic behaviour for avatars in Metaverse or in VR scenarios.
Integration
Customize the solution to your use case – send recognition results to third-party systems, integrate with virtual reality applications, or show results directly to the user.
3D Data Privacy and 3D data anonymization
We care about the privacy of your data. No matter whether 2D or 3D data, thanks to Edge AI implementation method, your data does not need to be sent to the cloud or other servers for processing, which guarantees data privacy and avoids data transfer problems.
Use our proprietary 3D anonymization technology if you want to anonymize and share the data with external partners without violating people’s privacy.
That's what we offer
Workshop and Use Case Analysis
We analyze your individual use case and conduct workshops to explain to you what are the capabilities of the technology to you.
Hardware and Setup Choice
According to your requirements we select the most optimal hardware and the best setup.
MVP
Building of Minimal Viable Product (MVP) to quickly verify the benefits of the technology for your use case.
Full Project
Implementation
We implement the whole project for you, without the need to build a large in-house team.
Support and AI Models Maintenance
We provide support and maintenance of the AI models to make sure that they continuously achieve great results.
Our exemplary implementations
Our deep tech expertise enables us to help you leverage the right technology to address your most complex and critical challenges. Read more about some of our successful 3D projects.
Seizure recognition
Automatic and contactless detection of complex seizure movements of epileptic patients.