Using Data Augmentation for Vision-Based Deep Reinforcement Learning

Dear Colleagues,

In recent years, vision-based deep reinforcement learning (RL) has emerged as a powerful approach for solving complex decision-making tasks by leveraging visual data. However, the performance and generalization of deep RL models can be significantly hampered by the limited availability of diverse and representative training data. This Special Issue focuses on the innovative application of data augmentation techniques to enhance vision-based deep RL. Data augmentation, which encompasses techniques such as image transformations, synthetic data generation, and domain randomisation, can enhance the robustness and versatility of models by artificially increasing the variability of training datasets. This Special Issue explores a wide range of augmentation strategies and their impact on the learning efficiency, stability, and generalisation capabilities of RL agents. Contributions include theoretical insights, novel algorithms, and practical applications across various domains. The aim of this Special Issue is to facilitate the advancement of research in the field of vision-based deep RL by addressing critical challenges and opening new avenues for investigation.

In this Special Issue, original research articles and reviews are welcome. Research areas may include (but are not limited to) the following:

• Techniques and methodologies for data augmentation in vision-based systems.
• Applications of RL in environments with visual input.
• Performance comparison of RL models with and without data augmentation.
• The impact of synthetic and real-world data on the learning efficiency and accuracy of RL systems.
• Case studies detailing the implementation of vision-based RL systems in various domains.
• Theoretical insights or reviews on the convergence properties of augmented RL algorithms.
• Innovations in hardware and software that enhance the training of RL systems using augmented data.

We look forward to receiving your contributions.

Dr. Kai Vahldiek
Prof. Dr. Dirk Joachim Lehmann
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Electronics is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2400 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

• data augmentation
• vision-based deep reinforcement learning
• image transformations
• synthetic data generation
• domain randomization
• learning efficiency
• generalization in reinforcement learning