Autonomous driving paper index
Optimizing AI Inference on Android Devices: A Comparative Study of GPU and NPU Acceleration
One-line summary
The pervasive integration of Artificial Intelligence (AI) into the Computing Continuum is driving a shift toward decentralized, real-time processing at the Edge.
Engineering notes
Our core objective is to empirically determine the combination that achieves the best trade-off between minimal accuracy degradation and maximal inference speed-up.
Chinese explanation / 中文解读
中文解读待补充：本站会优先为端到端自动驾驶、BEV感知、3D目标检测、轨迹预测、路径规划、LiDAR感知等高价值论文补充中文说明。
Original abstract
The pervasive integration of Artificial Intelligence (AI) into the Computing Continuum is driving a shift toward decentralized, real-time processing at the Edge. However, achieving autonomous and efficient Edge-Cloud operations requires overcoming significant challenges in performance sustainability and hardware heterogeneity. This paper investigates optimal execution strategies for AI models within the context of intelligent Edge service management, focusing on two critical tasks for infrastructure observability and smart services: object detection (YOLO family) and image classification (ResNet). These configurations evaluate various model quantization schemes and the utilization of on-device accelerators, specifically the GPU and NPU. Our core objective is to empirically determine the combination that achieves the best trade-off between minimal accuracy degradation and maximal inference speed-up.
Links and sources
Need this topic turned into a technical roadmap?
Full Self Driving can prepare a custom autonomous driving literature review, code map, dataset map, and B2B technology assessment.
Request B2B research
Comments