Publication: January-March 2025
The rapid advancement in AI technology continues to revolutionize multimedia applications, integrating them seamlessly into our daily activities. AI-powered multimedia applications have become indispensable, offering enhanced user experiences across various platforms. The development of foundation models holds great promise in further enhancing the inferability and reliability of multimedia data, driving innovation in this field.
Foundation models, characterized by their ability to be adapted to a multitude of downstream tasks, represent the next wave in AI evolution. These models are trained on extensive datasets using self-supervised learning techniques, enabling them to perform a wide range of functions across different domains. Despite their transformative potential, there remain significant challenges and opportunities in fully harnessing their capabilities for multimedia applications.
Active research topics in foundation models include AI agents, hallucination mitigation, long-context modeling, and robust automatic evaluation. AI agents leverage foundation models to perform complex tasks autonomously, while efforts to address hallucinations aim to enhance the accuracy and reliability of generated content. Long-context modeling focuses on improving the models’ ability to understand and utilize extensive contextual information, and robust automatic evaluation seeks to develop more reliable metrics for assessing model performance.
This special issue aims to bring together researchers and practitioners from both industry and academia to present the latest high-quality research and technical innovations in the field of foundation models for multimedia. By addressing current challenges and exploring new frontiers, we aim to advance the understanding and application of foundation models, paving the way for the next generation of AI-powered multimedia solutions.
Topics of interest include but are not limited to:
- Technical advances in foundation models (e.g., innovations in model architectures and training procedures)
- Applications of foundation models in industries (e.g., success stories and lessons learned from industrial deployments)
- Applications of foundation models in various areas (e.g., robotics and autonomous systems)
- The technical challenges and opportunities of foundation models (e.g., addressing issues of bias, fairness, and ethical considerations)
- Advances in technical principles behind foundation models (e.g., enhancements in model fine-tuning and adaptation techniques)
- Advanced AI systems involving foundation models (e.g., multimodal learning applications and cross-domain adaptability)
- Foundation models for next-generation intelligent services (e.g., future prospects and directions for intelligent service development)
- Other research topics that are closely related to foundation models for multimedia
Submission Guidelines
IEEE MultiMedia magazine seeks original articles discussing research and advanced practices in hardware and software, spanning the range from theory to working systems. We encourage our authors to write in a conversational style, presenting even technical material clearly and simply. You can use figures, tables, and sidebars to explain specific points, summarize results, define acronyms, guide readers to other sources, or highlight items. Assume an educated general audience, and you will successfully communicate your ideas to generalists and specialists alike.
Articles submitted to IEEE MultiMedia should not exceed 6,500 words, including all text, the abstract, keywords, bibliography, and biographies. Each table and figure counts for 200 words. Please limit the number of references to the 20 most relevant (except for surveys, which may have up to 30 references). The abstract should be no more than 150 words and should describe the overall focus of your manuscript.
Before submitting, please read our author guidelines. When you are ready to submit, please go to ScholarOne Manuscripts.
Questions?
Contact the guest editors at mm1-25@computer.org.
- Prof. Wen-Huang Cheng, National Taiwan University
- Prof. Jiaying Liu, Peking University
- Prof. Tanaya Guha, University of Glasgow
- Dr. Balu Adsumilli, YouTube/Google
- Dr. Kai-Lung Hua, Microsoft
- Dr. Yung-Hui Li, Foxconn Research
