Minimalist Visual Inertial Odometry

2026-05-19

AI Summary

This work proposes a minimalist planar odometry approach that overcomes the high computational demands of traditional visual-inertial systems, which rely on high-resolution cameras and are ill-suited for resource-constrained mobile platforms. The method employs only four downward-facing photodiodes equipped with Gabor optical masks, fused with an inertial measurement unit (IMU). By jointly optimizing the mask parameters and a temporal convolutional network (TCN), the system decodes linear velocity from the four-channel intensity signals and integrates IMU angular velocity to estimate trajectory. This study demonstrates, for the first time, that robust and accurate planar localization can be achieved with merely four visual sensing points. The approach closely tracks ground-truth trajectories across diverse indoor and outdoor terrains without requiring platform-specific fine-tuning, substantially reducing both sensor complexity and computational overhead.

Abstract

Visual-Inertial Odometry(VIO), which is critical to mobile robot navigation, uses cameras with a large number of pixels. Capturing and processing camera images requires significant resources. This work presents a minimalist approach to planar odometry, demonstrating that just four visual measurements and an IMU can provide robust motion estimation for differential-drive robots. Our key insight is that four downward-facing photodiodes that sense the world through optical Gabor masks produce signals that encode speed. Based on this, we jointly optimize the mask parameters alongside a Temporal Convolutional Network (TCN) using a physically-grounded simulator. The resulting model decodes speed from just the four measurements produced by the photodiodes. Pairing these estimates with the angular speed from an IMU yields a continuous planar trajectory. We validate our approach with a prototype sensor mounted on a differential drive robot. Across diverse indoor and outdoor terrains, our system closely tracks the reference ground truth without any real-world fine-tuning. Our work shows that minimalist sensing enables efficient and accurate planar odometry.

Problem

• Visual-Inertial Odometry
• minimalist sensing
• planar odometry
• resource efficiency
• motion estimation

Innovation

• Minimalist Sensing
• Visual-Inertial Odometry
• Gabor Masks
• Temporal Convolutional Network
• Planar Odometry