Exploring Domain Adaptation in Computer Vision

Dear Colleagues,

Humans inherently process visual information, effortlessly transferring knowledge across diverse visual scenarios. For instance, when presented with a novel visual situation, our innate cognition repurposes past experiences to decipher new challenges. Mirroring this human prowess, there is a growing focus on crafting computer vision algorithms that echo similar adaptability, harnessing the power of deep learning paradigms.

While traditional computer vision methods excel within their specific realms, they often face limitations due to their rigidity. These methods typically demand extensive retraining for every unique task, leading to computational inefficiencies. Moreover, they may overlook the potential synergies present among related visual tasks, such as image classification, object detection, and semantic segmentation.

Enter domain adaptation in computer vision—a pioneering approach designed to bridge these limitations. It champions the idea of transferring knowledge from one or multiple source domains to a pertinent target domain, mirroring human adaptability. The advent of unsupervised, semi-supervised, multi-source, and test-time domain adaptation techniques has equipped computer vision systems with the dexterity to tackle a range of tasks with limited supervision.

Cutting-edge techniques, especially in co-training, adversarial learning, and self-training, are injecting fresh impetus into this domain. This Special Issue delves deep into the latest strides in domain adaptation, shining a light on its myriad applications—from static image interpretation to dynamic video analysis. Our goal is to foster a comprehensive understanding, encouraging further research and collaboration in this rapidly evolving intersection of domain adaptation and computer vision.

Dr. Dayan Guan
Dr. Jiaxing Huang
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. Journal of Imaging 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 1800 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.

• unsupervised domain adaptation
• semi-supervised domain adaptation
• multi-source domain adaptation
• test-time domain adaptation
• image classification
• object detection
• semantic segmentation
• video analysis