Zero-shot Active Visual Search (ZAVIS): Intelligent Object Search for Robotic Assistants

TL;DR

This paper introduces AVSW, a flexible system enabling robots to locate objects described in free-form language by planning search paths based on semantic understanding and commonsense knowledge, validated in simulation and real-world tests.

Contribution

It presents a novel active visual search system that handles free-form language commands and plans efficient search paths using semantic maps and commonsense reasoning.

Findings

Outperforms previous methods in success-weighted path length in simulations.

Demonstrates effective real-world object search with a Pioneer-3AT robot.

Achieves higher success rate and efficiency in both simulated and real environments.

Abstract

In this paper, we focus on the problem of efficiently locating a target object described with free-form language using a mobile robot equipped with vision sensors (e.g., an RGBD camera). Conventional active visual search predefines a set of objects to search for, rendering these techniques restrictive in practice. To provide added flexibility in active visual searching, we propose a system where a user can enter target commands using free-form language; we call this system Active Visual Search in the Wild (AVSW). AVSW detects and plans to search for a target object inputted by a user through a semantic grid map represented by static landmarks (e.g., desk or bed). For efficient planning of object search patterns, AVSW considers commonsense knowledge-based co-occurrence and predictive uncertainty while deciding which landmarks to visit first. We validate the proposed method with respect to SR (success rate) and SPL (success weighted by path length) in both simulated and real-world environments. The proposed method outperforms previous methods in terms of SPL in simulated scenarios with an average gap of 0.283. We further demonstrate AVSW with a Pioneer-3AT robot in real-world studies.