ST Releases AI Feature Pack for Edge AI Vision Device Development
Published2021.02.23 10:25STM32 'FP-AI-VISION1' firmware feature pack and
Launch of the 'B-CAMS-OMV' Camera Hardware Bundle
STMicroelectronics today launched a new AI firmware function pack and camera module hardware bundle that enables embedded developers to implement computer vision applications that run locally at the edge using STM32 microcontrollers (MCUs).
Feature Pack and Camera Module Hardware Bundle Unveiled [Graphic = ST]
The 'FP-AI-VISION1' STM32Cube function pack, which can be easily ported to all STM32 MCUs, includes various code examples and implements a convolutional neural network (CNN) on the 'STM32H747' MCU. In addition to the various application examples suggested by the firmware, developers can retrain the neural network with their own dataset, giving them the freedom and flexibility to address a variety of use cases.
A new feature supports webcam mode for simple image collection using UVC (USB Video Camera). Code examples for food classification and human presence detection are available, and you can even implement a visual "wakeword" to wake the system from sleep mode. Developers can learn how to implement an image classification application using the Teachable Machine online tool, along with the STM32Cube.AI and FP-AI-VISION1 function packs, on the STM32 Wiki.
The 'B-CAMS-OMV' camera bundle is optimized for use with the FP-AI-VISION1 and provides the hardware required for training and deployment. It features the 5-megapixel ST 'MB1379 OV5640' color camera module, mounted on an adapter card compatible with all STM32 Discovery and evaluation boards with a ZIF connector. This adapter card can also be used with the ST 'VG5661' automotive HDR global shutter camera.
Additionally, Waveshare and OpenMV connectors allow connection to cameras of various wavelengths, enabling a wide range of computer vision applications. Instructions on how to integrate code generated with STM32Cube.AI into the OpenMV ecosystem can be found in the STM32 wiki content.
The FP-AI-VISION1 includes a variety of frame buffer processing functions, a camera driver, and software supporting image capture, preprocessing, and neural network inference. It provides a variety of neural network models, including quantized models and floating-point models generated with the neural network-optimized ST's "X-CUBE-AI" C code generator. Flexible memory configuration allows models to be tailored to the desired application.
The feature pack is available for free download from st.com. The B-CAMS-OMV camera module bundle is available for $56 from st.com and distributors.
관련뉴스
Cognex Unveils Color-Perceiving DL-Based Industrial Camera
딥러닝 기반 검사의 30%는 컬러 이미징이 필요하다. 색상은 여러 산업의 다양한 조립 검증 및 결함 감지 검사에서 특징을 구별 짓는 요소다. 이에 코그넥스가 컬러 이미징 기능을 새로 탑재한 딥러닝 기반 산업용 스마트 카메라인 인사이트 D900 컬러를 출시했다.
2020.11.06by 이수민 기자
Object recognition technology that improves autonomous driving safety, "How far has it come?"
자율주행차량은 카메라, 라이다, 레이더 등의 센서가 요구된다. 자율주행차량에 탑재된 객체 인식 AI 모델은 데이터셋을 기반으로 다수의 센서가 전달한 정보를 해석하여 주변 객체를 인지하고 방향과 구동을 결정한다. 객체 인식은 어느 분야든 레이블 된 데이터셋을 필요로 하지만, 자율주행 산업계는 아직 데이터셋 레이블 방식에 대한 합의를 이루지 못해 기초 표준만 나온 상태다.
2020.12.07by 이수민 기자
The "SensePro 2" DSP, offering six times the DSP performance and twice the AI capabilities, has been released.
CEVA has launched the SensePro2 DSP product family, which performs AI and DSP tasks related to a wide range of sensors including cameras, radar, and lidar. The new product family delivers 6 to 8 times higher DSP performance than SensePro1, 2 times higher AI inference performance, and 20% higher power efficiency on the same process node.
2021.01.20by 명세환 기자
ST Microelectronics and NUS Establish Joint Research Center 'HELIX' for Edge AI
ST Microelectronics and the National University of Singapore have established the 'HELIX' joint research center aimed at implementing embedded AI in edge environments. Over a four-year period, the center will conduct research across the entire technology stack from algorithms to silicon, focusing on memory-centric architecture leveraging ST's 18nm FD-SOI technology and embedded PCM. The core objective is to implement intelligence directly in physical devices such as robots and drones while minimizing energy consumption.
2026.08.28by 배종인 기자
ams, 디지털 포토닉스로 휴머노이드 로봇 감각·소통 기능 지원
ams OSRAM이 디지털 포토닉스 기술을 활용해 휴머노이드 로봇의 감각·소통 기능을 강화하는 솔루션을 공개했다. dToF 센서, 광학식 힘 센서, 구조광 이미터 등으로 주변 환경 인식과 정밀 조작을 구현하며, EVIYOS·ALIYOS 플랫폼으로 LED 기반 비언어 소통을 지원한다. 시장조사에 따르면 휴머노이드 로봇 시장은 2025년 6억 달러에서 2035년 510억 달러로 성장할 전망이며, 로봇의 감각 능력 향상이 핵심 성장 동인으로 작용할 것으로 예상된다.
2026.08.27by 배종인 기자
