Video doorbells? Machine vision? Autonomous mobile robots?
Texas Instruments had smart camera use cases like these in mind with its introduction of a new family of vision processors that, according to the company, will enable designers to easily and affordably add scalable edge AI performance for up to 12 cameras in building, industrial and retail automation applications.
A new family of vision processors is aimed at smart camera applications. Source: Texas InstrumentsThe family is comprised of a half-dozen Arm Cortex-based vision processors supported by open-source evaluation and model development tools, and common software that is programmable through industry-standard application programming interfaces (APIs), frameworks and models.
The processors feature system-on-a-chip (SoC) architecture that includes extensive integration. Integrated components include Arm Cortex-A53 or Cortex-A72 central processing units, a third-generation TI image signal processor, internal memory, interfaces, and hardware accelerators that deliver from 1 to 32 teraoperations per second (TOPS) of AI processing for deep learning algorithms.
The family includes:
- Four iterations of the AM62Ax: AM62A3, AM62A3-Q1, AM62A7 and AM62A7-Q1. With a maximum resolution of 5 MP, the AM62Ax models support one to two RGB-IR cameras at a power as low as 1 W, making them a fit for applications like video doorbells, smart retail systems, security cameras and vacuum robots. According to Texas Instruments, the AM62A3 is currently the industry’s lowest-cost 1-TOPS vision processor.
- AM68A, which is optimized for multi-inference real-time systems found in retail and factory automation. The processor enables one to eight RGB cameras in applications like machine vision, with up to 8 TOPS of AI processing for advanced video analytics. Maximum resolution is 12 MP; power can be as low as 6 W.
- AM69A, built for high-performance sensor fusion systems, achieves 32 TOPS of AI processing for one to 12 cameras in applications such as edge AI boxes, autonomous mobile robots and traffic monitoring systems. Maximum resolution 12 MP; power as low as 15 W.
TI has also announced that beginning in Q2 2023, designers can accelerate time to market for their edge AI applications with a public beta of a free open-source tool, Edge AI Studio — a feature-rich, web-based tool that allows users to develop and test AI models using user-created models and TI’s optimized models, which can also be retrained with custom data.
