Our project on human rendering from occluded videos addresses the challenge of accurately reconstructing 3D human bodies from single-view, in-the-wild video data, where unexpected occlusions often occur. By leveraging state-of-the-art generative AI, our methods recover unseen parts of the human body while maintaining both 3D structure and temporal consistency. This approach allows us to create realistic, continuous renderings of 3D humans, even when parts of the body are obscured, enhancing applications in virtual reality, animation, sports analysis, and healthcare.