Effects of Reward Terms in Agent-Based Box-Manipulation Animation Using Deep Reinforcement Learning

Yang, Hsiang-Yu; Wong, Chien-Chou; Wong, Sai‐Keung

doi:10.1109/taai48200.2019.8959860

Cited by 1 publication

(2 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For object transportation, there is a sparse reward problem in which a rare reward is received when a transportation task is completed. Thus, some methods rely on various reward terms (e.g., progress reward, time reward, and orientation reward) to guide agents to perform actions 9,15,17 . However, these methods require user intervention to adjust the weights of the reward terms.…”

Section: Related Workmentioning

confidence: 99%

“…Thus, some methods rely on various reward terms (e.g., progress reward, time reward, and orientation reward) to guide agents to perform actions. 9,15,17 However, these methods require user intervention to adjust the weights of the reward terms. To tackle the sparse reward problem with user intervention, an automatic curriculum learning can be employed, in which curricula are designed automatically for the agents during the training process.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Animation generation for object transportation with a rope using deep reinforcement learning

Wong

Wei

2023

Computer Animation & Virtual

View full text Add to dashboard Cite

This article presents a reinforcement learning‐based approach to generate animation in which two agents use a rope to collaboratively transport a block. The challenge is that the agents need to master several skills, including approaching the block, using the rope to wrap around it, and then moving the block to a predefined goal position. We propose several reward terms to learn the transportation policy and the adjustment policy that govern the skills of the agents. Experiment results showed that the proposed approach was able to generate various animations in different settings, including rope lengths, block sizes, and block shapes. An ablation test revealed the effects of the reward terms. We also investigated factors that affected the performance of the two policies.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%