# DVS-multi-cue-pedestrian-detection

**Repository Path**: hilbert-wang/DVS-multi-cue-pedestrian-detection

## Basic Information

- **Project Name**: DVS-multi-cue-pedestrian-detection
- **Description**: https://github.com/colinshane/DVS-multi-cue-pedestrian-detection
- **Primary Language**: Unknown
- **License**: MIT
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2024-10-23
- **Last Updated**: 2024-10-23

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Multi-Cue Event Information Fusion for Pedestrian Detection with Neuromorphic Vision Sensors

This repository contains the code for our paper (**in review**):

    Guang Chen, Hu Cao, Canbo Ye, Zhenyan Zhang, Xingbo Liu, Xuhui Mo, Zhongnan Qu, Jörg Conradt, Florian Röhrbein, Alois Knoll
    "Multi-Cue Event Information Fusion for Pedestrian Detection with Neuromorphic Vision Sensors"
    in Frontiers in Neurorobotics 2019

If you find the code useful for your research, please cite our paper (**in review**):

        @inproceedings{guangchen2019-dvs-ped-detection,
          title={Multi-Cue Event Information Fusion for Pedestrian Detection with Neuromorphic Vision Sensors},
          author={Guang Chen, Hu Cao, Canbo Ye, Zhenyan Zhang, Xingbo Liu, Xuhui Mo, Zhongnan Qu, Jörg Conradt, Florian Röhrbein, Alois Knoll},
          booktitle={Frontiers in Neurorobotics},
          year={2019}
        }

If you have any questions, please feel free to contact:

    Guang Chen <guangchen AT tongji.edu.cn>

## Introduction

We propose to develop pedestrian detectors that unlock the potential of the event data by leveraging multi-cue information and different fusion strategies.

To make the best out of the event data, we introduce three different event-stream encoding methods based on Frequency, Surface of Active Event (SAE) and Leaky Integrate-and-Fire (LIF). We further integrate them into the state-of-the-art neural network architectures with two fusion approaches: the channel-level fusion of the raw feature space and decision-level fusion with the probability assignments. We present a qualitative and quantitative explanation why different encoding methods are chosen to evaluate the pedestrian detection and which method performs the best.

## Algorithms

We promise that the code with algorithms involved in our paper will be made public here in this repository right after our paper is accepted and published.

## DVS-based Pedestrian Dataset

Based on our knowledge, no public labeled pedestrian dataset created with a neuromorphic vision sensor has already been published, so that we created one.

We promise that the dataset involved in our paper will be made public with a hyperlink here right after our paper is accepted and published.