Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Autobiasing Event Cameras for Flickering Mitigation

Mehdi Sefidgar Dilmaghani, Waseem Shariff, Cian Ryan, Joe Lemley, Peter Corcoran

TL;DR

This work tackles flicker in event cameras by deploying an autonomous autobiasing system that tunes the bias_fo to suppress flicker across $25$-$500$ Hz. It employs a CNN-based flicker detector integrated into a feedback loop that adjusts a sensor bias in real time, tested within a GR-YOLO/YOLO V3 face-detection framework. The approach uses Average Gradient as a universal flicker metric and app-specific metrics (YOLO confidences, face-detection success) to quantify improvements. Results show significant reductions in AG and substantial gains in face-detection performance under both well-lit and low-light conditions, indicating that single-bias tuning can robustly mitigate flicker across diverse environments and frequencies.

Abstract

Understanding and mitigating flicker effects caused by rapid variations in light intensity is critical for enhancing the performance of event cameras in diverse environments. This paper introduces an innovative autonomous mechanism for tuning the biases of event cameras, effectively addressing flicker across a wide frequency range -25 Hz to 500 Hz. Unlike traditional methods that rely on additional hardware or software for flicker filtering, our approach leverages the event cameras inherent bias settings. Utilizing a simple Convolutional Neural Networks -CNNs, the system identifies instances of flicker in a spatial space and dynamically adjusts specific biases to minimize its impact. The efficacy of this autobiasing system was robustly tested using a face detector framework under both well-lit and low-light conditions, as well as across various frequencies. The results demonstrated significant improvements: enhanced YOLO confidence metrics for face detection, and an increased percentage of frames capturing detected faces. Moreover, the average gradient, which serves as an indicator of flicker presence through edge detection, decreased by 38.2 percent in well-lit conditions and by 53.6 percent in low-light conditions. These findings underscore the potential of our approach to significantly improve the functionality of event cameras in a range of adverse lighting scenarios.

Autobiasing Event Cameras for Flickering Mitigation

TL;DR

This work tackles flicker in event cameras by deploying an autonomous autobiasing system that tunes the bias_fo to suppress flicker across - Hz. It employs a CNN-based flicker detector integrated into a feedback loop that adjusts a sensor bias in real time, tested within a GR-YOLO/YOLO V3 face-detection framework. The approach uses Average Gradient as a universal flicker metric and app-specific metrics (YOLO confidences, face-detection success) to quantify improvements. Results show significant reductions in AG and substantial gains in face-detection performance under both well-lit and low-light conditions, indicating that single-bias tuning can robustly mitigate flicker across diverse environments and frequencies.

Abstract

Understanding and mitigating flicker effects caused by rapid variations in light intensity is critical for enhancing the performance of event cameras in diverse environments. This paper introduces an innovative autonomous mechanism for tuning the biases of event cameras, effectively addressing flicker across a wide frequency range -25 Hz to 500 Hz. Unlike traditional methods that rely on additional hardware or software for flicker filtering, our approach leverages the event cameras inherent bias settings. Utilizing a simple Convolutional Neural Networks -CNNs, the system identifies instances of flicker in a spatial space and dynamically adjusts specific biases to minimize its impact. The efficacy of this autobiasing system was robustly tested using a face detector framework under both well-lit and low-light conditions, as well as across various frequencies. The results demonstrated significant improvements: enhanced YOLO confidence metrics for face detection, and an increased percentage of frames capturing detected faces. Moreover, the average gradient, which serves as an indicator of flicker presence through edge detection, decreased by 38.2 percent in well-lit conditions and by 53.6 percent in low-light conditions. These findings underscore the potential of our approach to significantly improve the functionality of event cameras in a range of adverse lighting scenarios.

Paper Structure

This paper contains 21 sections, 2 equations, 7 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Event Cameras Hardware and Biases
  3. bias_fo (buffer bandwidth)
  4. Literature Review
  5. Flicker Mitigation in Frame-based Cameras
  6. Flickering in Event Camera
  7. Event Camera Tuning and Biases
  8. Methodology
  9. Events Processor
  10. Event-based Application
  11. Efficacy Metrics Calculation Block
  12. Average Gradient (AG)
  13. Application Specific Assessment Metrics
  14. Flickering Detection Classifier
  15. Network Architecture
  16. ...and 6 more sections

Figures (7)

  • Figure 1: The hardware of each pixel in an event camera m10.
  • Figure 2: Auto-biasing block diagram
  • Figure 3: Architecture of the flickering classification CNN
  • Figure 4: Samples of frames: a) without flickering, b) with flickering
  • Figure 5: Bias control and optimization process: example of face detection
  • ...and 2 more figures