[2024 ICME Workshop] Vision and Learning for an Enhanced Metaverse

Jun LIU
Singapore University of Technology and Design

Hossein RAHMANI
Lancaster University

Mengyuan LIU
Peking University

Ping HU
University of Electronic Science and Technology of China

Zhipeng FAN
Meta

Kian Eng ONG
Singapore University of Technology and Design

Xun Long NG
Singapore University of Technology and Design

Tony QUEK
Singapore University of Technology and Design

METAVERSE

... the next evolution in social connection ... (Meta)

... the next iteration of the internet that seamlessly combines our digital and physical lives ... (McKinsey)


Avatar 19 July 2024

VLMetaverse Workshop


Programme Schedule
2pm to 5pm, Peninsular 1F
Zoom meeting link
Meeting ID: 857 3686 1809
Passcode: 349226
  • Keynote Address: Shaping Virtual Realities: Diffusion Models for Metaverse Applications by Dr Lin Geng Foo, Singapore University of Technology and Design
  • Oral presentation
    • A Survey on Backbones for Deep Video Action Recognition
    • High-Fidelity 3D Model Generation with Relightable Appearance from Single Freehand Sketches and Text Guidance
Shaping Virtual Realities: Diffusion Models for Metaverse Applications
by Dr Lin Geng Foo, Singapore University of Technology and Design

Diffusion models have emerged as powerful tools for generative AI in enriching metaverse experiences with interactive and diverse content. These models excel in generating varied and customizable outputs, hence making them well-suited for creating the vast amounts of detailed digital assets needed for immersive virtual worlds. Besides the generative tasks, diffusion models can be applied to non-generative tasks that are important for enabling the metaverse. This includes 3D pose estimation, 3D mesh recovery, and action detection. However, it is challenging to directly apply the usual method of diffusion to these tasks. As such, in this talk, he will share about how to tackle these challenges.



Avatar 22 February 2024

Call for Papers



Click on the respective headers to view / collapse details
The Metaverse is an immersive and interactive multimedia platform that needs to impeccably integrate sound, image, mesh, video, animation, and text to create a seamless user experience in the virtual world. This brings about numerous opportunities and also diverse challenges.

Hence, this workshop aims to bring together and unite researchers and practitioners (e.g., animators, game developers, data engineers) in

  1. Scene understanding, person and/or object recognition for the Metaverse
  2. Generative AI for scene, person and/or object generation or reconstruction for the Metaverse
  3. Emotion detection, social interaction, and/or human-machine interactions in the Metaverse
  4. Computationally efficient storage, compression, transmission, and/or processing of data, models, animations etc for the Metaverse
  5. Challenges and solutions in addressing concerns about privacy, safety, intellectual property, and/or security of data and models for Metaverse applications
so that we can all create a more advanced and secure Metaverse accessible to all.

The workshop will consist of talks by Keynote Speakers, as well as oral and poster presentations by authors of accepted papers.

We invite high-quality papers that are potentially useful, related or can be applied to the Metaverse (i.e., not necessarily have already been implemented in the Metaverse), including but not limited to the following themes:
  1. 2D / 3D / Video / Multi-modal scene understanding, person and/or object recognition for the Metaverse
    • Algorithms for scene understanding, person and/or object recognition (including object-object interactions) in the Metaverse.

    • These shall facilitate the models to learn about a scene, person, and/or object that is logical and beneficial in the Metaverse. For example, object and scene understanding algorithms allow the reconstruction of real-life scenes and logical object interactions in the Metaverse.

  2. Generative AI for scene, person and/or object generation or reconstruction for the Metaverse
    • Algorithms for scene, person, and/or object generation, including 2D to 3D reconstruction (e.g., NeRF).

    • These shall be useful to generate or reconstruct realistic-looking 3D objects in the Metaverse from real-life scenes, objects, or images. For example, this can be useful for fashion or e-commerce companies to enable users to see how the generated clothes or fashion accessories fit them.

  3. Emotion detection, social interaction, and/or human-machine interactions in the Metaverse
    • Algorithms for emotion detection and social interactions, including pose/gestures and facial recognition, in the Metaverse.

    • These shall facilitate the development of AI systems that are more human-like and are able to interpret emotions and respond appropriately to users. This can also include novel ways for human-machine interactions. For example, gesture recognition algorithms enable users to logically interact with objects and other people in the Metaverse.

  4. Computationally efficient storage, compression, transmission, and/or processing of data, models, animations etc for the Metaverse
    • Computationally efficient AI solutions for Metaverse, including storage, compression, transmission and/or processing of data, models, images, animations or outputs in cloud / edge computing.

    • These will help enhance the user experience or maximize interactions between individuals while minimizing latency. These shall also be useful conserving energy and resources for a better and sustainable Earth as various industry players embark on the Metaverse.

  5. Challenges and solutions in addressing concerns about privacy, safety, intellectual property, and/or security of data and models for Metaverse applications
    • Algorithms for privacy-preserving (including intellectual property and security), safe, or unbiased models for Metaverse applications (e.g., consideration of marginalized and vulnerable populations).

    • These ethical, privacy, and social issues are growing in importance and greater awareness about them needs to be raised and be subsequently tackled by various domain experts (including AI engineers, cybersecurity engineers, network engineers, cloud developers, and app developers).
We also accept survey or meta-review papers that are relevant or useful for the Metaverse.


Avatar 31 March 2024

Submission Guidelines



Types of Paper Submission

  • Each paper will be evaluated based on the following criteria:
    • (Potential) relevance to the Metaverse (i.e., is relevant / can potentially be useful or implemented in the Metaverse)
    • Potential impact (if applicable, including novelty of idea, ethical considerations behind findings of the paper etc) on the field
    • Quality of writing
  • All accepted papers will be presented as either oral or poster presentations. The accepted papers will appear in ICME proceedings.
  • The Best Paper Award will be awarded to the most insightful / useful / impactful research work.
  • A workshop paper is covered by a full-conference registration only.

Paper Submission

  • The paper will undergo a double-blind review, which means that authors cannot know the names of the reviewers of their papers, and reviewers cannot know the names of the authors. Information that may identify the authors anywhere in the submitted materials must be avoided. In particular, in the submitted pdf paper, the usual list of authors, their institutions, and their contact information must be replaced by the phrase, “Anonymous ICME Submission.” Identifying information in the acknowledgments (e.g., co-workers and grant IDs), supplemental materials (e.g., titles in the videos, or attached papers), and links to the authors’ or their institutions’ websites must also be avoided. The only place that authors may be identified is in the online submission form.
  • You should explicitly highlight in the abstract and introduction and/or discussions about how your work (or meta-review) is relevant or can potentially be useful or potentially implemented in the Metaverse.
  • The paper submission will follow the ICME 2024 Workshops guidelines (https://2024.ieeeicme.org/author-information-and-submission-instructions)
  • Length: Papers must be no longer than 6 pages, including all text, figures, and references.
  • Format: Workshop papers have the same format as regular papers.
  • Submission site: Papers must be submitted under the corresponding track under ICME 2024 Workshop-VLMetaverse. Submissions may be accompanied by up to 20 MB of supplemental material following the same guidelines as regular and special session papers.
  • Presentation guarantee: As with accepted Regular and Special Session papers, accepted Workshop papers must be registered by the author deadline and presented at the conference; otherwise they will not be included in IEEE Xplore.
  • A workshop paper is covered by a full-conference registration only.


Avatar July 2024

Workshop Schedule



The workshop will consist of talks by Keynote Speakers, as well as oral and poster presentations by authors of accepted papers.

More details to come soon.


31 March 2024
UTC 23:59:00

Deadline for submitting papers (Select: ICME 2024 Workshop-VLMetaverse)

11 April 2024
UTC 23:59:00

Notification of paper acceptance

01 May 2024
Deadline for camera-ready submission of accepted paper

01 May 2024
Author Full Conference Registration