Transformative and Secure Computer Vision Devices

IoT devices that use innovative computer vision technologies are transforming the world around us and delivering new experiences in diverse environments—from retail to smart homes.


These computer vision technologies are improving how cities optimize resources, increase mobility, and provide health and safety benefits. They’re also transforming manufacturing through improved production-line monitoring and smart robotics.


To bring these optimizations and opportunities to life, vision-enabled devices rely on increased compute capabilities, pulling together imaging, machine learning, security, and cloud enablement into one heterogenous system.

Collaboration and Innovation on Arm

The Arm ecosystem is advancing the development of vision hardware, software, and systems with smart vision solutions across all market sectors.

 

Plumerai: People Detection on Embedded Devices with Arm Helium Technology

Plumerai showcases computer vision solutions on small, embedded devices to address use cases that include people detection and face identification. The platform is built on the high-performance Renesas RA8x1 MCU with Arm Cortex-M85 and Arm Helium vector extensions to accelerate people-detection neural networks using just 300 KB RAM.

Himax: Enabling AI Inferencing on Resources-Constrained Edge Devices

Himax WiseEye 2 runs vision-based AI edge models, including face detection, human pose estimation, and object detection. Its heterogeneous solution combines Arm Cortex-M55 with Arm Helium technology, the Arm Ethos-U55 NPU, and support for TensorFlow Lite Micro and TVM, to enable high-speed vision processing.

Grovety: Object Detection - Saving Wildlife with AI-powered Cameras

Grovety reveals how it solved real-life object detection challenges, including battery and power constraints, false positives, and limited data availability. By combining Arm Cortex-A32, Arm Cortex-M55, and Arm Ethos-U55 NPU, while leveraging the TVM and Vela compilers, it created an autonomous AI-powered device for object detection and classification, adaptable to wildlife objects.

A Strong Ecosystem for Smart Vision

Arm partners form a strong, collaborative ecosystem for intelligent vision designs on Arm and cover hardware, software, and system integrations—all ready to address the breadth of vision applications.

Unlock the Power of Smart Vision

Arm is advancing the development of vision-enabled devices with the hardware IP, solutions, software, tools, and standards to achieve performance and scalability with accelerated time – to market.

IP and Solution icon

IP and Solutions

With over 80% of smart cameras today built on Arm-based technology, we have expanded on this experience to provide a portfolio of products that meet the needs of the breadth of vision devices. This portfolio spans the CPU, ISP, NPU, and GPU, with IP that offers choice and flexibility for the wide-ranging needs of vision platforms.

Software icon

Software

Arm offers a rich ecosystem of supported software to streamline and accelerate development. Arm-NN, CMSIS-NN, and Vela Compiler simplify the development of ML applications and integrations with third-party software stacks, to reduce software complexities.

Tools icon

Tools

Arm offers tools to support development. Arm Keil and Arm Development Studio accelerate design and development, while the developer experience is further optimized by tools such as Arm IP Explorer and Arm Virtual Hardware. These help simplify the process of IP selection and enable software development ahead of hardware availability.

Standards

Strong foundational standards, such as SystemReady, PSA Certified, and CMSIS, ensure that portability and security are built into every device to help speed development and innovation at scale.

How to Build Computer Vision Solutions

How to Build Computer Vision Solutions

Devices that can see and act on visual information have a significant advantage for IoT in terms of functionality and efficiency. However, designing and deploying a competitive end-to-end solution for smart vision is a complicated task, often beyond the expertise of one company. This paper outlines the factors developers should consider before starting a design and when facing the challenges of integrating the necessary IP, software, tools, and standards for smart vision.

Download White Paper