Belief Tree Search for Active Object Recognition

TL;DR

This paper introduces a belief tree search approach for active object recognition, modeling it as a POMDP, and demonstrates improved recognition accuracy and policy optimization through supervised learning and observation function refinement.

Contribution

The paper presents a novel POMDP-based framework for AOR, using belief tree search and LSTM prediction, with a gradient-based method to optimize the observation function.

Findings

LSTM-based policy outperforms guided policy search in recognition accuracy.

Optimizing the observation function improves total reward and recognition performance.

The approach generalizes well to new views and objects.

Abstract

Active Object Recognition (AOR) has been approached as an unsupervised learning problem, in which optimal trajectories for object inspection are not known and are to be discovered by reducing label uncertainty measures or training with reinforcement learning. Such approaches have no guarantees of the quality of their solution. In this paper, we treat AOR as a Partially Observable Markov Decision Process (POMDP) and find near-optimal policies on training data using Belief Tree Search (BTS) on the corresponding belief Markov Decision Process (MDP). AOR then reduces to the problem of knowledge transfer from near-optimal policies on training set to the test set. We train a Long Short Term Memory (LSTM) network to predict the best next action on the training set rollouts. We sho that the proposed AOR method generalizes well to novel views of familiar objects and also to novel objects. We compare this supervised scheme against guided policy search, and find that the LSTM network reaches higher recognition accuracy compared to the guided policy method. We further look into optimizing the observation function to increase the total collected reward of optimal policy. In AOR, the observation function is known only approximately. We propose a gradient-based method update to this approximate observation function to increase the total reward of any policy. We show that by optimizing the observation function and retraining the supervised LSTM network, the AOR performance on the test set improves significantly.