In modern cyber-physical systems, the integration of AI into vision pipelines is now a standard practice for applications ranging from autonomous vehicles to mobile devices. Traditional AI integration often relies on cloud-based processing, which faces challenges such as data access bottlenecks, increased latency, and high power consumption. This article reviews embedded AI vision systems, examining the diverse landscape of near-sensor and in-sensor processing architectures that incorporate convolutional neural networks. We begin with a comprehensive analysis of the critical characteristics and metrics that define the performance of AI-integrated vision systems. These include sensor resolution, frame rate, data bandwidth, computational throughput, latency, power efficiency, and overall system scalability. Understanding these metrics provides a foundation for evaluating how different embedded processing architectures impact the entire vision pipeline, from image capture to AI inference. Our analysis delves into near-sensor systems that leverage dedicated hardware accelerators and commercially available components to efficiently process data close to their source, minimizing data transfer overhead and latency. These systems offer a balance between flexibility and performance, allowing for real-time processing in constrained environments. In addition, we explore in-sensor processing solutions that integrate computational capabilities directly into the sensor. This approach addresses the rigorous demand constraints of embedded applications by significantly reducing data movement and power consumption while also enabling in-sensor feature extraction, pre-processing, and CNN inference. By comparing these approaches, we identify trade-offs related to flexibility, power consumption, and computational performance. Ultimately, this article provides insights into the evolving landscape of embedded AI vision systems and suggests new research directions for the development of next-generation machine vision systems.
7 January 2019 in IEEE Journal of Solid-State Circuits
This paper presents a 3-D stacked vision chip featuring in-focal-plane read-out tightly coupled with flexible computing architecture for configurable high-speed image analysis. The chip architecture is based on a scalable standalone structure integrating image sensor on the top tier and processing elements (PEs) plus memories in the bottom tier. By using 3-D stacking partitioning, our prototype benefits from backside illuminated pixels sensitivity, a fully parallel communication between image sensor and PEs for low-latency performances, while leaving enough room in the bottom tier to embed advanced computing features. One scalable structure embeds a 16x16 pixel array (or 64x64 pixels in high-resolution mode), associated with an 8-bit single instruction multiple data (SIMD) processor; fabricated in dual 130-nm 1P6M CMOS process. This paper exhibits a 5500 frames/s and 85 giga operations per second (GOPS)/W in low-resolution mode, with large kernels capabilities through eight directions interpixel communication. Multiflow capability is also demonstrated to execute different programs in different areas of the vision chip.
Three-dimensional-integrated focal-plane array image processor chips offer new opportunities to implement highly parallelised computer vision algorithms directly inside sensors. Neural networks in particular can perform highly complex machine vision tasks, and therefore their efficient implementation in such imagers are of significant interest. However, studies with existing pixel-processor array chips have focused on the implementation of a subset of neural network components—notably convolutional kernels—on pixel processor arrays. In this work, we implement a continuous end-to-end pipeline for a convolutional neural network from the digitisation of incoming photons to the output prediction vector on a macropixel processor array chip (where a single processor acts on group of pixels). Our implementation performs inference at a rate between 265 and 309 frames per second, directly inside of the sensor, by exploiting the different levels of parallelism available.