Activity, Plan, and Goal Recognition: A Review

Abstract

Recognizing the actions, plans, and goals of a person in an unconstrained environment is a key feature that future robotic systems will need in order to achieve a natural human-machine interaction. Indeed, we humans are constantly understanding and predicting the actions and goals of others, which allows us to interact in intuitive and safe ways. While action and plan recognition are tasks that humans perform naturally and with little effort, they are still an unresolved problem from the point of view of artificial intelligence. The immense variety of possible actions and plans that may be encountered in an unconstrained environment makes current approaches be far from human-like performance. In addition, while very different types of algorithms have been proposed to tackle the problem of activity, plan, and goal (intention) recognition, these tend to focus in only one part of the problem (e.g., action recognition), and techniques that address the problem as a whole have been not so thoroughly explored. This review is meant to provide a general view of the problem of activity, plan, and goal recognition as a whole. It presents a description of the problem, both from the human perspective and from the computational perspective, and proposes a classification of the main types of approaches that have been proposed to address it (logic-based, classical machine learning, deep learning, and brain-inspired), together with a description and comparison of the classes. This general view of the problem can help on the identification of research gaps, and may also provide inspiration for the development of new approaches that address the problem in a unified way.

Two ways to make your robot proactive: Reasoning about human intentions or reasoning about possible futures

“…Approaches in action recognition, goal recognition, and plan recognition have been used to infer intention. According to Van-Horenbeke and Peer (2021) , intention recognition systems can be classified as logic-based, classical machine learning, deep learning, and brain-inspired approaches, or they can be classified in terms of the behavior of the observed human toward the observer. We take here a simplified view and consider two classes of intention recognition approaches: logic-based and probabilistic.…”

Comparing Plan Recognition Algorithms Through Standard Plan Libraries

“…Previous papers surveyed methods for plan recognition, but did not try to run and evaluate the presented algorithms as well (Carberry, 2001;Sukthankar et al, 2014;Albrecht and Stone, 2017;Mirsky et al, 2021;Van-Horenbeke and Peer, 2021). Many of these surveys discuss the research efforts in plan, activity and intent recognition (PAIR) and present open challenges that are relevant to the PAIR community, such as: 1) designing computationally efficient algorithms, and 2) finding a single representation that can naturally generate various modelling capabilities.…”

Towards Computational Modeling of Human Goal Recognition

“…A comprehensive survey of the field, which has started in the late 1970s (Schmidt et al, 1978), is well beyond the scope of this paper. We refer the interested party to several comprehensive recent surveys by Sukthankar, Goldman, Geib, Pynadath and Bui (Sukthankar et al, 2014), by Mirsky, Keren, and Geib (Mirsky et al, 2021), and by Van-Horenbeke and Peer (Van-Horenbeke and Peer, 2021), as well as to an earlier survey by Carberry (Carberry, 2001).…”