FROG: A new people detection dataset for knee-high 2D range finders

2023-06-14

arXiv.org

AI Summary

Existing 2D lidar-based human detection datasets for knee-height scanning (e.g., DROW) suffer from low spatial resolution, insufficient frame rate, and incomplete annotations, limiting their applicability to real-time perception for mobile robots operating from low vantage points. To address these limitations, we introduce FROG—a new benchmark dataset featuring 100% fully annotated lidar frames, significantly higher spatial resolution, and increased temporal sampling frequency. We further propose an end-to-end deep learning model based on a CNN-LSTM hybrid architecture that directly processes raw point sequences without handcrafted features or domain-specific priors. The system is optimized for real-time deployment within ROS, achieving inference speeds exceeding 500 Hz. FROG scales to 17× the total scan volume and 100× the number of human annotations compared to DROW, while attaining state-of-the-art detection performance.

Abstract

Mobile robots require knowledge of the environment, especially of humans located in its vicinity. While the most common approaches for detecting humans involve computer vision, an often overlooked hardware feature of robots for people detection are their 2D range finders. These were originally intended for obstacle avoidance and mapping/SLAM tasks. In most robots, they are conveniently located at a height approximately between the ankle and the knee, so they can be used for detecting people too, and with a larger field of view and depth resolution compared to cameras. In this paper, we present a new dataset for people detection using knee-high 2D range finders called FROG. This dataset has greater laser resolution, scanning frequency, and more complete annotation data compared to existing datasets such as DROW. Particularly, the FROG dataset contains annotations for 100% of its laser scans (unlike DROW which only annotates 5%), 17x more annotated scans, 100x more people annotations, and over twice the distance traveled by the robot. We propose a benchmark based on the FROG dataset, and analyze a collection of state-of-the-art people detectors based on 2D range finder data. We also propose and evaluate a new end-to-end deep learning approach for people detection. Our solution works with the raw sensor data directly (not needing hand-crafted input data features), thus avoiding CPU preprocessing and releasing the developer of understanding specific domain heuristics. Experimental results show how the proposed people detector attains results comparable to the state of the art, while an optimized implementation for ROS can operate at more than 500 Hz.

Problem

• Detecting humans using knee-high 2D range finders
• Improving dataset quality for people detection benchmarks
• Developing efficient deep learning for raw sensor data

Innovation

• New dataset FROG for knee-high 2D range finders
• End-to-end deep learning with raw sensor data
• ROS-optimized detector operates over 500 Hz