The field of 4D augmented reality is rapidly advancing, with a focus on improving the fidelity and coherence of 4D representations. Researchers are exploring new methods for generating and representing 4D content, including the use of deep learning models and novel frameworks for 4D generation. These advancements have the potential to enable more immersive and interactive experiences in virtual reality, animation, and other applications. A key challenge in this field is addressing the limitations of current hardware and communication systems, which can restrict the complexity and detail of 4D models. To overcome this, researchers are developing compact representations of 4D video sequences and optimizing 4D Gaussians for dynamic scene video. Notable papers in this area include:

• Video4DGen, which presents a novel framework for generating 4D representations from single or multiple generated videos.
• Uni4D, which introduces a unified self-supervised learning framework for point cloud videos, achieving state-of-the-art performance on several benchmarks.