Machine Learning Techniques for Computer Vision

Dear Colleagues,

Computer vision can be used to enhance the reliability of communications through wireless networks and to improve human–computer interaction. With the rapid development of machine learning techniques, research in computer vision has made significant progress in recent years. Deep learning methods have brought revolutionized breakthroughs in many computer vision tasks. Techniques have been developed to process stationary images and the combination of video and audio input.

However, the “black box” nature of the deep learning model blocks its application to computer vision tasks for high-stakes decision making. Developing interpretable deep learning models is thus desired in this application area. Many post hoc interpretability analysis methods have been developed to understand pre-trained models. Ad hoc interpretability modeling methods have also been developed, such as feature disentanglement and interpretable model extraction using mimic learning.

Similar to large language models, transformer-based large language models have also been intensively used for tackling computer vision tasks. However, due to the high cost of data labeling in special application scenarios, self-supervised learning, few-shot learning, transfer learning, or other new machine learning techniques must be developed to address the lack of labeled data.

In addition, applications in autonomous vehicles, augmented reality, and smart cities are driving the evolution of internet infrastructure for computer vision in profound ways. Autonomous vehicles require real-time data processing for navigation and decision making, relying on robust internet connectivity for updates and synchronization. Augmented reality applications demand high-bandwidth connections to stream immersive content seamlessly. Smart city initiatives leverage computer vision for traffic management, surveillance, and infrastructure monitoring, necessitating scalable networks to handle the influx of data. These applications are catalyzing advancements in internet infrastructure to support the growing demands of computer vision technologies in various domains.

The goal of this Special Issue is to provide a platform for researchers to share their new findings and ideas in the area of tackling computer vision tasks with the development of machine learning methods. Submissions may include reviews, surveys, or technical papers that are original and unpublished, with topic areas including, but not limited to, the following:

• Computer vision for human–computer interaction;
• Image processing on mobile devices;
• Self-supervised learning for computer vision;
• Few-shot learning for computer vision;
• Transfer learning for computer vision;
• Interpretable machine learning methods for computer vision;
• Machine learning methods for image classification;
• Machine learning methods for object detection;
• Machine learning methods for semantic segmentation;
• Machine learning methods for video processing;
• Machine learning methods for action recognition;
• Machine learning methods for anomaly detection;
• Autonomous vehicles and computer vision;
• Augmented reality and computer vision;
• Smart cities and computer vision.

Prof. Dr. Yonggang Lu
Dr. Ivan Serina
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. Future Internet is an international peer-reviewed open access monthly 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 1600 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.

• machine learning
• computer vision
• mobile devices
• human–computer interaction
• deep learning
• image processing
• video processing
• few-shot learning
• transfer learning
• interpretability
• image classification
• object detection
• semantic segmentation
• action recognition
• anomaly detection