Image analysis-based artificial intelligence (AI) models leveraging convolutional neural networks (CNN) take a significant role in evaluating the ripeness of strawberry, contributing to the maximization of productivity. However, the convolution, which constitutes the majority of the CNN models, imposes significant computational burdens. Additionally, the dense operations in the fully connected (FC) layer necessitate a vast number of parameters and entail extensive external memory access. Therefore, reducing the computational burden of convolution operations and alleviating memory overhead is essential in embedded environment. In this paper, we propose a strawberry ripeness classification system utilizing a convolution-based feature extractor (CoFEx) for accelerating convolution operations and an edge AI processor, Intellino, for replacing FC layer operations. We accelerated feature map extraction utilizing the CoFEx constructed with systolic array (SA) and alleviated the computational burden and memory overhead associated with the FC layer operations by replacing them with the k-nearest neighbors (k-NN) algorithm. The CoFEx and the Intellino both were designed with Verilog HDL and implemented on a field-programmable gate array (FPGA). The proposed system achieved a high precision of 93.4%, recall of 93.3%, and F1 score of 0.933. Therefore, we demonstrated a feasibility of the strawberry ripeness classification system operating in an embedded environment.
14 March 2024 in IEEE Transactions on Circuits and Systems II: Express Briefs
Designing high-performance hardware sorter for resource-constrained systems is challenging due to physical limitations and the need to balance streaming bandwidth with memory throughput. This paper introduces a novel, scalable hardware sorter architecture with fully-streaming support and an accompanying RTL generator to provide versatile, energy-efficient hardware acceleration. Our solution employs a dual-layer architecture consisting of a parallel one-way linear insertion sorter (OLIS) for bandwidth optimization and a cyclic bitonic merge network (CBMN) for a compact, high-throughput implementation. Furthermore, we developed the RTL generator written in Chisel to provide the agile implementation of the scalable architecture. Experimental results targeting the Xilinx XVU37P-FSVH2892-2L-E FPGA show that our design achieves throughput increasing by 126.26% and latency decreasing by 68.46%, with an area increment of no more than 132.94% for LUTs and a decrement of flip-flops by 79.84%, compared to state-of-the-art streaming sorter. The source code is available at https://github.com/hyun-woo-oh/DL-Sort-Generator.
Autonomous robots operate on batteries, rendering power efficiency essential. The substantial computational demands of object detection present a significant burden to the low-power cores employed in these robots. Therefore, we propose a grid-based density-based spatial clustering of applications with a noise (DBSCAN) clustering accelerator for light detection and ranging (LiDAR)’s point cloud to accelerate computational speed and alleviate the operational burden on low-power cores. The proposed method for DBSCAN clustering leverages the characteristics of LiDAR. LiDAR has fixed positions where light is emitted, and the number of points measured per frame is also fixed. These characteristics make it possible to impose grid-based DBSCAN on clustering a LiDAR’s point cloud, mapping the positions and indices where light is emitted to a 2D grid. The designed accelerator with the proposed method lowers the time complexity from to . The designed accelerator was implemented on a field programmable gate array (FPGA) and verified by comparing clustering results, speeds, and power consumption across various devices. The implemented accelerator speeded up clustering speeds by 9.54 and 51.57 times compared to the i7-12700 and Raspberry Pi 4, respectively, and recorded a 99% reduction in power consumption compared to the Raspberry Pi 4. Comparisons of clustering results also confirmed that the proposed algorithm performed clustering with high visual similarity. Therefore, the proposed accelerator with a low-power core successfully accelerated speed, reduced power consumption, and effectively conducted clustering.
Transformer-based models have achieved remarkable success across various AI tasks, but their growing complexity has led to significant computational and memory demands. While most optimization efforts have focused on linear operations such as matrix multiplications, non-linear functions like Softmax and layer normalization (LayerNorm) are increasingly dominating inference latency, especially for long sequences and high-dimensional inputs. To address this emerging bottleneck, we present a hardware accelerator that jointly approximates these non-linear functions using piecewise linear approximation for the exponential in Softmax and Newton–Raphson iteration for the square root in LayerNorm. The proposed unified architecture dynamically switches operation modes while reusing hardware resources. The proposed accelerator was implemented on a Xilinx VU37P FPGA and evaluated with BERT and GPT-2 models. Experimental results demonstrate speedups of up to 7.6× for Softmax and 2.0× for LayerNorm, while maintaining less than 1% accuracy degradation on classification tasks with conservative approximation settings. However, generation tasks showed greater sensitivity to approximation, underscoring the need for task-specific tuning.
Currently, the trend of elderly people living alone is rising due to rapid aging and shifts in family structures. Accordingly, the efficient implementation and management of monitoring systems tailored for elderly people living alone have become paramount. Monitoring systems are generally implemented based on multiple sensors, and the collected data are processed on a server to provide monitoring services to users. Due to the use of multiple sensors and a reliance on servers, there are limitations to economical maintenance and a risk of highly personal information being leaked. In this paper, we propose an intelligent monitoring system with privacy preservation based on edge AI. The proposed system achieves cost competitiveness and ensures high security by blocking communication between the camera module and the server with an edge AI module. Additionally, applying edge computing technology allows for the efficient processing of data traffic. The edge AI module was designed with Verilog HDL and was implemented on a field-programmable gate array (FPGA). Through experiments conducted on 6144 frames, we achieved 95.34% accuracy. Synthesis results in a 180 nm CMOS technology indicated a gate count of 1516 K and a power consumption of 344.44 mW.
Convolution operation is a crucial technique in the field of artificial intelligence (AI), particularly in image processing-based applications. However, a significant amount of computation time is required to perform an operation on a vast amount of data. The conventional operation processing method of the systolic array architecture tends to accelerate the speed of convolution operations by reusing only the weight data. To minimize data movement time, we propose a systolic array architecture that partially reuses the input feature map. Compared to the conventional systolic array accelerator, the proposed architecture demonstrated a throughput improvement of ×6.