Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

HAPQ: A Hardware-Aware Pruning and Quantization Pipeline for Event-Based SNN Detection.

Zhengyinan Li1, Jing Wu1,2

  • 1School of Automotive Engineering, Wuhan University of Technology, Wuhan 430070, China.

Sensors (Basel, Switzerland)
|May 13, 2026
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

Downsampling01:20

Downsampling

When considering a sampled sequence with zero values between sampling instants, one can replace it by taking every N-th value of the sequence. At these integer multiples of N, the original and sampled sequences coincide. This process, known as decimation, involves extracting every N-th sample from a sequence, thereby creating a more efficient sequence.
The Fourier transform of the decimated sequence reveals a combination of scaled and shifted versions of the original spectrum. This...
Upsampling01:22

Upsampling

Managing signal sampling rates is essential in digital signal processing to maintain signal integrity. A decimated signal, characterized by a reduced frequency range due to its lower sampling rate, can be upsampled by inserting zeros between each sample. This upsampling process expands the original spectrum and introduces repeated spectral replicas at intervals dictated by the new Nyquist frequency. To refine this zero-inserted sequence, it is passed through a lowpass filter with a cutoff...

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

[The role of endoplasmic reticulum stress in insulin resistance of type 2 diabetes].

Sheng li ke xue jin zhan [Progress in physiology]·2012
Same author

EVI1 acts as an inducible negative-feedback regulator of NF-κB by inhibiting p65 acetylation.

Journal of immunology (Baltimore, Md. : 1950)·2012
Same author

Cu(II)-catalyzed asymmetric hydrosilylation of diaryl- and aryl heteroaryl ketones: application in the enantioselective synthesis of orphenadrine and neobenodine.

Chemistry (Weinheim an der Bergstrasse, Germany)·2012
Same author

T-bet acts as a powerful adjuvant in Ag85B DNA‑based vaccination against tuberculosis.

Molecular medicine reports·2012
Same author

Aspirin-triggered lipoxin A₄ attenuates lipopolysaccharide-induced intracellular ROS in BV2 microglia cells by inhibiting the function of NADPH oxidase.

Neurochemical research·2012
Same author

Preparation and time-gated luminescence bioimaging applications of long wavelength-excited silica-encapsulated europium nanoparticles.

Nanoscale·2012
Same journal

RETRACTED: Zhang et al. A Novel Framework for Reconstruction and Imaging of Target Scattering Centers via Wide-Angle Incidence in Radar Networks. <i>Sensors</i> 2025, <i>25</i>, 6802.

Sensors (Basel, Switzerland)·2026
Same journal

Enhancing Unsupervised Multi-Source Domain Adaptation for Person Re-Identification via Mixture of Experts and Graph-Based Relation.

Sensors (Basel, Switzerland)·2026
Same journal

Development of an Instrumented Glove for Palmar Pressure Assessment in Kayakers.

Sensors (Basel, Switzerland)·2026
Same journal

Development and Experimental Validation of an Autonomous IoT-Based Monitoring System for Real-Time Water Quality Assessment in the Amazon River.

Sensors (Basel, Switzerland)·2026
Same journal

Semi-Supervised Adversarial Learning Framework for Controller Area Network Bus Intrusion Detection.

Sensors (Basel, Switzerland)·2026
Same journal

Smart Optimization Method for Safety Signs in Innovative Manufacturing Environments Integrating Industrial Field IoT Sensors and Knowledge Graphs.

Sensors (Basel, Switzerland)·2026
See all related articles

This study introduces HAPQ, a novel pipeline optimizing Spiking Neural Networks (SNNs) for event-based object detection on FPGAs. It significantly boosts accuracy and efficiency for autonomous driving perception systems.

Area of Science:

  • Computer Vision
  • Artificial Intelligence
  • Hardware Acceleration

Background:

  • Autonomous driving perception requires low latency, high temporal resolution, and hardware efficiency.
  • Event-based Spiking Neural Networks (SNNs) offer sparse computation but face challenges in FPGA deployment due to irregular patterns and state storage.
  • Existing methods struggle to balance performance and resource constraints for SNNs on edge devices.

Purpose of the Study:

  • To develop a unified hardware-aware pruning and quantization pipeline (HAPQ) for compact event-based object detection.
  • To enable efficient deployment of SNNs on Field-Programmable Gate Arrays (FPGAs) for autonomous driving.
  • To address the limitations of irregular execution and temporal state storage in SNNs.

Main Methods:

Keywords:
FPGAevent-based object detectionhardware-aware co-designmembrane potential quantizationstructured pruning

Related Experiment Videos

  • Proposing HAPQ, a pipeline integrating hardware-aware configuration search, SIMD-aligned structured pruning, and joint quantization of weights and potentials.
  • Utilizing an end-to-end adaptive sampling SNN detector (EAS-SNN) as the base model.
  • Evaluating on the Prophesee Gen1 dataset and an FPGA accelerator, optimizing within Digital Signal Processor (DSP) and Block RAM (BRAM) budgets.

    • Main Results:

    • HAPQ improved mean average precision (mAP50:95) from 0.284 to 0.425, achieving 0.722 mAP50.
    • Hardware implementation showed reduced Lookup Table (LUT) usage to 1680, eliminated DSPs, and operated at 920.81 MHz with 0.630 W power consumption.
    • Demonstrated significant reductions in resource utilization and power consumption while enhancing detection accuracy.

    Conclusions:

    • HAPQ effectively enables compact event-based object detection for autonomous driving on FPGAs.
    • Joint optimization of SNN architecture, state precision, and hardware-aligned workload organization is crucial for efficient temporal SNN deployment.
    • The proposed method offers a viable solution for deploying high-performance SNNs on resource-constrained edge devices.