Efficient Learning for Computer Vision: Few-Shot, Weakly Supervised and Unsupervised Approaches

Dear colleagues,

Recent advances in computer vision have been propelled by the advent of deep learning and the availability of large labeled datasets. However, its dependence on an extensive amount of labeled data presents a significant barrier, as such data are often costly and time-consuming to acquire and impractical in many domains. This Special Issue addresses this challenge by focusing on efficient learning methods, including few-shot, weakly supervised, and unsupervised learning. These approaches aim to reduce the reliance on large labeled datasets, making computer vision more versatile and widely applicable. Few-shot learning enables models to generalize from a minimal number of examples, weakly supervised learning leverages coarse annotations for tasks requiring fine supervision, and unsupervised learning extracts meaningful patterns from data without any labels. These methodologies are critical for expanding the application of computer vision into areas like medical imaging, remote sensing, and autonomous systems where labeled data are scarce or expensive.

The aim of this Special Issue is to present cutting-edge research that advances the development of efficient learning in computer vision. It will highlight innovative techniques, algorithms, and applications in relation to few-shot, weakly supervised, and unsupervised learning, demonstrating these methods’ potential to tackle complex vision tasks with limited supervision. This aligns with the scope of the journal, which showcases pioneering developments in computer vision and machine learning, particularly those addressing data efficiency.

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

• Few-shot learning for tasks such as image classification, object detection, and semantic segmentation.
• Meta-learning algorithms for rapid adaptation to new visual tasks.
• Weakly supervised methods for dense prediction tasks like segmentation and localization.
• Unsupervised representation learning for visual data.
• Self-supervised learning techniques utilizing pretext tasks.
• The use of generative models for data augmentation in few-shot scenarios.
• Efficient architectures designed for learning with limited data.
• Real-world applications, such as in healthcare, autonomous driving, or robotics.
• The theoretical foundations of efficient learning in computer vision.
• Benchmarking and evaluation metrics for these learning paradigms.

We look forward to receiving your valuable contributions.

Dr. Lien Minh Dang
Prof. Dr. Hyeonjoon Moon
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.

• computer vision
• few-shot learning
• weakly supervised learning
• unsupervised learning
• meta-learning
• self-supervised learning
• object detection
• image segmentation