The central focus of this journal is the computer analysis of pictorial information. Computer Vision and Image Understanding publishes papers covering all aspects of image analysis from the low-level, iconic processes of early vision to the high-level, symbolic processes of recognition and interpretation. A wide range of topics in the image understanding area is covered, including papers offering insights that differ from predominant views.

Research Areas Include:

• Theory
• Early vision
• Data structures and representations
• Shape
• Range
• Motion
• Matching and recognition
• Architecture and languages
• Vision systems

Special Issue on Eyes on People: Recent Trends on Human Analysis, Perception and Generation
Submission Date: 2024-03-21

Human-centered data are extremely widespread and have been intensely investigated by researchers belonging to very different fields. These research efforts are motivated by the several highly-informative aspects of humans that can be investigated, ranging from corporal elements to emotions and outward appearance. Recent advancements in machine and deep learning opened the way to the practical implementation of methods that can effectively blend into the real world, becoming for all intents and purposes an integral part of our everyday lives. The span of possible applications is vast, ranging from security and prevention of crimes to healthcare, social behavior, and even entertainment.
On the other hand, the extreme variety and diffusion of human-centered data, and the unprecedented level of accuracy of learning methods make their analysis and use extremely challenging and critical, as it normally involves personal and sensitive information. The improper use of such human-centered technology can have unpredictable social consequences if not properly taken into account. Indeed, the collection and use of human-centered data to build AI models capable of analyzing and predicting human behaviors raise non-negligible concerns related to privacy, ethics, and fairness.To conclude, we believe research efforts should go in a direction where learning models meet the always stricter privacy and ethics requirements. Overall, the goal of this Special Issue is to gather research efforts in the field of human analysis, perception, and generation. We aim at improving the communication between researchers and companies, and developing novel ideas that can shape the future of this area, in terms of motivations, methodologies, prospective trends, ethics, and potential industrial applications.

Guest editors:
Guido Borghi, PhDUniversity of Bologna, Bologna, Italy
Andrea Pilzer, PhDNVIDIA, Italy
Mohamed Daoudi, PhDIMT Nord Europe, France
Xavier Alameda Pineda, PhDINRIA Grenoble, France
Federico Becattini, PhDUniversity of Siena, Siena, Italy
Marcella Cornia, PhDUniversity of Modena and Reggio Emilia, Modena, Italy
Claudio Ferrari, PhDUniversity of Parma, Parma, Italy
Tomaso Fontanini, PhDUniversity of Parma, Parma, Italy

Special Issue on Advances in Deep Learning for Human-Centric Visual Understanding
Submission Date: 2024-04-30

Our daily lives revolve around people. One of artificial intelligence's primary goals is to create intelligent machines that enable humans to accomplish more and to live better lives. This requires machines to comprehend people’s emotional and physical characteristics, behaviors, and daily activities, among other things. As a result, human-centric visual comprehension is a critical and long-standing area of research in computer vision and artificial intelligence. It has a plethora of critical applications in our society, including security and safety, health care, and human-machine interfaces. Recent advances in deep learning have led to efficient and effective tools for dealing with the variability and complexity inherent in real-world environments. While significant progress has been made, there is still a significant gap in order to address complex human-centric visual reasoning tasks (e.g., understanding human-object interaction, analyzing human body language) and new challenges (e.g., face forgery detection). Thus, now is an excellent time to refocus research efforts on more comprehensive and in-depth human-centric visual comprehension, and ultimately on socially intelligent machines.
We welcome submissions of high-quality papers that introduce significant new theories, methods, applications, and insights into a variety of human-centric perception, reasoning, and analysis tasks. Possible subjects include, but are not limited to:

Human semantic parsing/fashion recognition
Human pose/shape estimation
Human activity recognition and trajectory prediction
Face detection/facial landmark detection/deepfake detection
Pedestrian detection/tracking/recognition/retrieval/re-identification
Human-object/-human interaction understanding
Human gaze/facial/body behavior analysis
Human visual attention mechanisms
Human-centric image/video synthesis
New benchmark datasets and survey papers related to the aforementioned topics

Guest editors:
Wenguan Wang, PhDZhejiang University, Hangzhou, China
Si Liu, PhDBeihang University, Beijing, China
Xiaojun Chang, PhDUniversity of Technology Sydney, Sydney, Australia
David Crandall, PhDIndiana University, Bloomington, United States of America
Haibin Ling, PhDStony Brook University, Stony Brook, United States of America 

Special Issue on Trustworthy Cross-Modal Reasoning for Video-Language Understanding
Submission Date: 2024-04-30

Video-language understanding and reasoning are long-standing problems for the CV and Multimedia communities. By endowing an AI machine with the crossmodality reasoning ability for video-language understanding, AI researchers expect the machine to “think” like a human and then make trustable decisions. However, most existing efforts primarily aim to improve in-domain performance while overlooking how to truly capture the essence of cross-modal reasoning. Especially the fundamental question in video-language understanding (Whether the model simply learns multimodal correlations hidden in datasets and whether it yields reliable in-domain results?) is usually overlooked by researchers and has yet to be well answered. Therefore, this special issue covers the continual growth of research, primarily related to the robustness, fairness, explainability, and security of video-oriented cross-modal reasoning.

The purpose of this special issue is to solicit high-quality, high-impact, and original papers on current developments in cross-modal reasoning for video-language understanding. We are interested in submissions covering topics of particular interest that include but are not limited to the following:

New datasets for trustworthy video-language understanding
Adversarial learning for robust multimodal representation
New methods for robust video summarization
Cross-modal semantics-consistent representation learning
Domain generalization in video-language understanding
Causal learning for trustworthy multimodal reasoning
Unfair bias measurement and mitigation in video-language understanding
Explainable multimodal data fusion and interaction
Brain-inspired networks for explainable cross-modal reasoning
Trustworthy reasoning algorithm in video dialog
Knowledge-driven explainable cross-modal reasoning
Text-guided visual-textual reasoning and generation
Privacy protection and security control in cross-modal AIGC
Applications of trustworthy video-language understanding

Guest editors:
Dan Guo, PhDHefei University of Technology, Hefei, China
Zhun Zhong, PhDUniversity of Nottingham, Nottingham, United Kingdom
Subhankar Roy, PhDTélécom Paris, Paris, France,
Linchao Zhu, PhDZhejiang University, Hangzhou, China, 
Chuang Gan, PhDUMass Amherst, Amherst, United States of AmericaMIT-IBM Watson AI Lab, Cambridge, United States of America
Meng Wang, PhDHefei University of Technology, Hefei, China