Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

HOOD: Real-Time Human Presence and Out-of-Distribution Detection Using FMCW Radar

Sabri Mustafa Kahya, Muhammet Sami Yavuz, Eckehard Steinbach

TL;DR

This work proposes a robust and real-time capable human presence and out-of-distribution (OOD) detection method using 60-GHz short-range FMCW radar that outperforms state-of-the-art OOD detection methods in terms of common OOD detection metrics.

Abstract

Detecting human presence indoors with millimeter-wave frequency-modulated continuous-wave (FMCW) radar faces challenges from both moving and stationary clutter. This work proposes a robust and real-time capable human presence and out-of-distribution (OOD) detection method using 60 GHz short-range FMCW radar. HOOD solves the human presence and OOD detection problems simultaneously in a single pipeline. Our solution relies on a reconstruction-based architecture and works with radar macro and micro range-Doppler images (RDIs). HOOD aims to accurately detect the presence of humans in the presence or absence of moving and stationary disturbers. Since HOOD is also an OOD detector, it aims to detect moving or stationary clutters as OOD in humans' absence and predicts the current scene's output as "no presence." HOOD performs well in diverse scenarios, demonstrating its effectiveness across different human activities and situations. On our dataset collected with a 60 GHz short-range FMCW radar, we achieve an average AUROC of 94.36%. Additionally, our extensive evaluations and experiments demonstrate that HOOD outperforms state-of-the-art (SOTA) OOD detection methods in terms of common OOD detection metrics. Importantly, HOOD also perfectly fits on Raspberry Pi 3B+ with an ARM Cortex-A53 CPU, which showcases its versatility across different hardware environments. Videos of our human presence detection experiments are available at: https://muskahya.github.io/HOOD

HOOD: Real-Time Human Presence and Out-of-Distribution Detection Using FMCW Radar

TL;DR

This work proposes a robust and real-time capable human presence and out-of-distribution (OOD) detection method using 60-GHz short-range FMCW radar that outperforms state-of-the-art OOD detection methods in terms of common OOD detection metrics.

Abstract

Detecting human presence indoors with millimeter-wave frequency-modulated continuous-wave (FMCW) radar faces challenges from both moving and stationary clutter. This work proposes a robust and real-time capable human presence and out-of-distribution (OOD) detection method using 60 GHz short-range FMCW radar. HOOD solves the human presence and OOD detection problems simultaneously in a single pipeline. Our solution relies on a reconstruction-based architecture and works with radar macro and micro range-Doppler images (RDIs). HOOD aims to accurately detect the presence of humans in the presence or absence of moving and stationary disturbers. Since HOOD is also an OOD detector, it aims to detect moving or stationary clutters as OOD in humans' absence and predicts the current scene's output as "no presence." HOOD performs well in diverse scenarios, demonstrating its effectiveness across different human activities and situations. On our dataset collected with a 60 GHz short-range FMCW radar, we achieve an average AUROC of 94.36%. Additionally, our extensive evaluations and experiments demonstrate that HOOD outperforms state-of-the-art (SOTA) OOD detection methods in terms of common OOD detection metrics. Importantly, HOOD also perfectly fits on Raspberry Pi 3B+ with an ARM Cortex-A53 CPU, which showcases its versatility across different hardware environments. Videos of our human presence detection experiments are available at: https://muskahya.github.io/HOOD
Paper Structure (12 sections, 2 equations, 2 figures, 3 tables, 1 algorithm)

This paper contains 12 sections, 2 equations, 2 figures, 3 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. RADAR SYSTEM DESIGN
  4. Pre-processing
  5. E-RESPD
  6. Proposed Method
  7. Architecture and Training
  8. Human and OOD Detection
  9. Experiments
  10. Dataset and Evaluation Metrics
  11. Conducted Experiments
  12. Conclusion

Figures (2)

  • Figure 1: This figure presents the high-level structure of HOOD (top) and provides a detailed view of one encoder-decoder pair within the architecture (bottom). Here the input data are radar macro and micro RDIs from static and very-static categories. The RDIs are processed through the network to extract meaningful representations. The architecture follows a reconstruction-based approach to reconstruct and interpret the input RDIs. In the high-level structure, the top and bottom encoders share the same architecture and are responsible for encoding macro and micro RDIs, respectively. The four decoders, in top-to-bottom order, decode macro RDIs for static activities, macro RDIs for very-static activities, micro RDIs for static activities, and micro RDIs for very-static activities, also having identical structures.
  • Figure 2: Illustration of the working mechanism of E-RESPD for pre-processing of macro and micro RDIs. E-RESPD utilizes a sliding-window approach. The window size is set to 200 frames, equivalent to 10 seconds of data acquisition from the radar sensor. The process starts with the first 200 frames being summed and written back to the first frame. Then, the window is shifted by one, and the next 200 frames in the window are summed and written back to the second frame. This process continues until the end of the data.