Call for Papers: Traditional Computer Vision in the Age of Deep Learning
- Matteo Poggi, University of Bologna, firstname.lastname@example.org
- Federica Arrigoni, University of Trento, email@example.com
- Andrea Fusiello, University of Udine, firstname.lastname@example.org
- Stefano Mattoccia, University of Bologna, email@example.com
- Adrien Bartoli, Université Clermont Auvergne, firstname.lastname@example.org
- Torsten Sattler, Czech Technical University in Prague, email@example.com
- Tomas Pajdla, Czech Technical University in Prague, firstname.lastname@example.org
In the last 5-10 years we have witnessed that deep learning has revolutionized Computer Vision, conquering the main scene in most vision journals/conferences. However, a number of problems and topics for which deep-learned solutions are currently not preferable over classical ones exist, that typically involve a strong mathematical model (e.g., camera calibration, structure from motion, ... ). The goal is to provide a venue where these topics (that might be overshadowed by the large number of problems tackled with deep learning) can get more attention. This special issue also aims at encouraging lines of research that might be considered out of fashion (although not out of date) by most scholars, such as theoretical results and geometric aspects of computer vision. Moreover, the special issue will encourage critical discussions about whether preferring a traditional solution rather than a deep learning approach, and it will also explore relevant questions about how to bridge the gap between learning and classic knowledge.
This special issue concentrates on algorithms and methodologies that address Computer Vision problems in a “traditional/classic” way, in the sense that analytical/explicit models are deployed, as opposed to learned/neural ones. Particular focus will be given to traditional approaches that perform better than neural ones (for instance, in terms of generalization across different domains) or that, although performing sub-par, provide clear advantages with respect to deep learning solutions (for instance, in terms of efforts to collect data, computational requirements, power consumption and more). All the Computer Vision topics are welcome as long as the proposed method does not solely consist in training an end-to-end neural model. Neural networks are not excluded, though: the special issue welcomes learning solutions that exploit traditional computer vision as their core (e.g., via geometric constraints) or hybrid approaches that combine deep networks with classical pipelines (e.g., by using learned local features in a structure from motion pipeline).
Submission deadline: January 31, 2022
First review notification (tentative): April 30, 2022
Revision due: May 31, 2022
Final decision: July 31, 2022
Submitted papers should present original, unpublished work, relevant to one of the topics of the Special Issue. All submitted papers will be evaluated on the basis of relevance, significance of contribution, technical quality, scholarship, and quality of presentation, by at least three independent reviewers. It is the policy of the journal that no submission, or substantially overlapping submission, be published or be under review at another journal or conference at any time during the review process. Manuscripts will be subject to a peer reviewing process and must conform to the author guide lines available on the IJCV website at: https://www.springer.com/11263 .
Please select "Traditional Computer Vision in the Age of Deep Learning" at the beginning of the submission process.
Authors are encouraged to submit high-quality, original work that has neither appeared in, nor is under consideration by other journals.
All papers will be reviewed following standard reviewing procedures for the Journal.
Papers must be prepared in accordance with the Journal guidelines: www.springer.com/11263
Other links include: