Accelerating Reinforcement Learning using EEG-based implicit human feedback

Xu, Duo; Agarwal, Mohit; Gupta, Ekansh; Fekri, Faramarz; Sivakumar, R.

doi:10.1016/j.neucom.2021.06.064

Cited by 22 publications

(9 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…When comparing the BGS tasks, BGSInt performs the best for both EEGNet (g * =0.77, CI.95[0.47, 1.08]) with a medium to large ES and SVM (g * =0.56, CI.95[0.26, 0.86]) with a small to large ES. Similarly, when comparing the OA tasks, OAOut performs best for EEGNet (g * =−1.82, CI.95[−2.16, −1.47]) 8 and SVM (g * =−1.66, CI.95[−2.0, −1.32]) with large ESs. Similar results to bACC can be observed for TNR (Fig.…”

Section: Resultsmentioning

confidence: 94%

“…ErrPs are generally characterized by a negative peak, often referred to as error-related negativity (ERN), that occurs between 80-300 ms and a subsequent positive peak (Pe) that occurs between 200-500 ms [1], [2], [4]- [6]. Recently, ErrPs have been of particular interest for BCI applications, ranging from simple detection/classification of errors [5]- [7], to training robots and agents [8], [9], to correcting mistakes made by BCI-controlled interfaces [10].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Error-related Potential Variability: Exploring the Effects on Classification and Transferability

Poole¹,

Lee²

2023

Preprint

View full text Add to dashboard Cite

Brain-Computer Interfaces (BCI) have allowed for direct communication from the brain to external applications for the automatic detection of cognitive processes such as error recognition. Error-related potentials (ErrPs) are a particular brain signal elicited when one commits or observes an erroneous event. However, due to the noisy properties of the brain and recording devices, ErrPs vary from instance to instance as they are combined with an assortment of other brain signals, biological noise, and external noise, making the classification of ErrPs a non-trivial problem. Recent works have revealed particular cognitive processes such as awareness, embodiment, and predictability that contribute to ErrP variations. In this paper, we explore the performance of classifier transferability when trained on different ErrP variation datasets generated by varying the levels of awareness and embodiment for a given task. In particular, we look at transference between observational and interactive ErrP categories when elicited by similar and differing tasks. Our empirical results provide an exploratory analysis into the ErrP transferability problem from a data perspective.1 Execution ErrPs are elicited with continuous actions while interaction ErrPs are elicited with discrete actions [6]

show abstract

Section: Resultsmentioning

confidence: 94%

Section: Introductionmentioning

confidence: 99%

Error-related Potential Variability: Exploring the Effects on Classification and Transferability

Poole¹,

Lee²

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…ErrP has been widely used in human-in-the-loop RL systems. In these systems, the ErrP signal is used as a positive or negative reward to accelerate the training of autonomous agents [31][32][33]. In [19], ErrPs was used as negative reinforcers of the actions to infer the optimal control strategies.…”

Section: Errp-based Human-in-the-loop Rlmentioning

confidence: 99%

“…In [2], ErrP was used to train a robot to learn human gestures through a reinforcement learning strategy based on the leap motion and ErrP features. However, when the ErrP signal was used as a reward, while the ErrPs accelerated learning, the signals operated independently of the system during testing [31][32][33].…”

Section: Errp-based Human-in-the-loop Rlmentioning

confidence: 99%

See 1 more Smart Citation

Error-related potential-based shared autonomy via deep recurrent reinforcement learning

Wang¹,

Chen²,

Lin³

2022

J. Neural Eng.

View full text Add to dashboard Cite

Error-related potential (ErrP)-based brain-computer interfaces (BCIs) have received a considerable amount of attention in the human-robot interaction community. In contrast to traditional BCI, which requires continuous and explicit commands from an operator, ErrP-based BCI leverages the ErrP, which is evoked when an operator observes unexpected behaviours from the robot counterpart. This paper proposes a novel shared autonomy model for ErrP-based human-robot interaction. We incorporate ErrP information provided by a user as useful observations for an agent and formulate the shared autonomy problem as a partially observable Markov decision process (POMDP). A recurrent neural network-based actor-critic model is used to address the uncertainty in the ErrP signal. We evaluate the proposed framework using a simulated human-in-the-loop robot navigation scenario with both simulated users and real users. The results show that the proposed ErrP-based shared autonomy model enables an autonomous robot to complete navigation tasks more efficiently. In a simulation with 70\% ErrP accuracy, agents completed the task 14.1% faster than in the no ErrP condition, while with real users, agents completed the navigation task 14.9 % faster.

show abstract

Understanding emotional values of bionic features for educational service robots: A cross-age examination using multi-modal data

Wang,

Li,

Shi

et al. 2024

Advanced Engineering Informatics

View full text Add to dashboard Cite

Accelerating Reinforcement Learning using EEG-based implicit human feedback

Cited by 22 publications

References 33 publications

Error-related Potential Variability: Exploring the Effects on Classification and Transferability

Error-related Potential Variability: Exploring the Effects on Classification and Transferability

Error-related potential-based shared autonomy via deep recurrent reinforcement learning

Understanding emotional values of bionic features for educational service robots: A cross-age examination using multi-modal data

Contact Info

Product

Resources

About