See the world in a whole new way with our AI camera technology
Safety AI solution is a type of video surveillance system that uses artificial intelligence (AI) technology to analyze and interpret live video feeds from CCTV cameras. They are designed to improve the efficiency and accuracy of video surveillance and provide real-time alerts and insights to security personnel. The solution has PPE detection, fall detection, safety harness detection, de-identification, and territory intrusion features. They offer benefits ranging from real-time alerts to behavioral analytics and are increasingly used in a variety of environments, including warehouses, construction sites, hospitals, and smart factories.
Action recognition
Analyzes video data and identifies specific actions performed by individuals or groups. Detect a variety of actions, from simple gestures to complex movements. Areas of development include slip and fall detection, property protection and medical monitoring.
Tracking and identification
Analyze video footage and track objects in real time. It can identify and track specific individuals or objects based on their appearance, behavior, or other characteristics. Developments include security, traffic monitoring, and retail analytics.
Lightweighting and optimization
Lightweight and optimize AI models on edge devices. Reduce model size to improve efficiency, save memory to improve inference speed, and enable models to mimic behavior to work more effectively without sacrificing performance. Deploy on resource-constrained edge devices.
Synthetic data
Synthetic data technology refers to the process of creating artificial data that mimics real data. It is used in machine learning and artificial intelligence applications to generate large amounts of data for training algorithms. One of the main benefits of synthetic data is that it can be used to train machine learning models without the need for large amounts of real data, saving time on labeling. This is especially useful in places like industry and healthcare, where real user data is sensitive and difficult to obtain. We are creating data that is focused on people and can provide a variety of locations, weather, and labeling types include bounding box, segmentation, human keypoints, depth, motion vector and etc.
Hardware modules
HW Modules offers custom embedded PCBs and NVIDIA development board. The PCB design must take into account factors such as the type of components needed, the power requirements of the device, and the board layout. Once the design is complete, the PCB is fabricated and a development environment is set up. It is designed to meet the unique requirements of each device, ensuring the performance and reliability of the device. We build PCBs based on our own SoCs and third-party SoCs. You can find carrier boards for NVIDIA Xavier NX and Orin NX, as well as Qualcomm-based mobile AI devices.
Miro AI chip
Design a 32-bit RISC-V-based microcontroller for low-cost AI inference applications. We determined that RISC-V, rather than ARM, was a more appropriate specification for the microcontroller market. The combination of TinyML and AIoT enables the development of intelligent connected devices that perform complex tasks while minimizing power consumption and latency. It is suitable for companies that need gesture recognition, number plate recognition, keyword extraction, collision avoidance, smart toys, human and face recognition, and absentee recognition. The 28nm process was used for competitive and technical reasons.
