Efficient Spatio-Temporal Signal Recognition on Edge Devices Using PointLCA-Net

2024-11-21

arXiv.org

AI Summary

To address the challenges of real-time processing, limited memory, and constrained power budgets for spatiotemporal signal recognition on edge devices, this paper proposes PointLCA-Net—a lightweight and energy-efficient architecture. Our method pioneers the integration of PointNet-based feature extraction with a spiking local competitive algorithm (LCA) for encoding–decoding, synergistically coupled with an in-memory computing paradigm for point cloud features implemented on memristor crossbar arrays to co-optimize training and inference. Compared to state-of-the-art approaches, PointLCA-Net achieves high recognition accuracy while significantly reducing both inference and training energy consumption, enabling real-time, low-latency, and ultra-low-power spatiotemporal signal processing at the edge. The core innovation lies in the co-design of point cloud in-memory computation and spiking LCA, which breaks the longstanding trade-off between energy efficiency and accuracy in edge intelligence.

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Cloud, edge and content delivery systems for the Web
• Search and Retrieval-Augmented AI: Large language models for search
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs

Abstract

Recent advancements in machine learning, particularly through deep learning architectures like PointNet, have transformed the processing of three-dimensional (3D) point clouds, significantly improving 3D object classification and segmentation tasks. While 3D point clouds provide detailed spatial information, spatio-temporal signals introduce a dynamic element that accounts for changes over time. However, applying deep learning techniques to spatio-temporal signals and deploying them on edge devices presents challenges, including real-time processing, memory capacity, and power consumption. To address these issues, this paper presents a novel approach that combines PointNet's feature extraction with the in-memory computing capabilities and energy efficiency of neuromorphic systems for spatio-temporal signal recognition. The proposed method consists of a two-stage process: in the first stage, PointNet extracts features from the spatio-temporal signals, which are then stored in non-volatile memristor crossbar arrays. In the second stage, these features are processed by a single-layer spiking neural encoder-decoder that employs the Locally Competitive Algorithm (LCA) for efficient encoding and classification. This work integrates the strengths of both PointNet and LCA, enhancing computational efficiency and energy performance on edge devices. PointLCA-Net achieves high recognition accuracy for spatio-temporal data with substantially lower energy burden during both inference and training than comparable approaches, thus advancing the deployment of advanced neural architectures in energy-constrained environments.

Problem

• Efficient spatio-temporal signal recognition on edge devices
• Challenges in real-time processing, memory, and power consumption
• Integration of PointNet and LCA for energy-efficient classification

Innovation

• Combines PointNet with neuromorphic computing
• Uses memristor crossbar arrays for storage
• Employs LCA for efficient encoding and classification