Casual3DHDR: High Dynamic Range 3D Gaussian Splatting from Casually Captured Videos

Shucheng Gong*, Lingzhe Zhao*, Wenpu Li*, Hong Xie, Yin Zhang, Shiyu Zhao, Peidong Liu

Acknowledgement: Xiang Liu

Project ArXiv Code

• We present Casual3DHDR, the first method for casual 3D high dynamic range (HDR) imaging using hand-held devices. Our method reconstructs high-quality 3D HDR scenes from a set of low dynamic range (LDR) images captured with varying exposure settings, addressing challenges such as dynamic scenes, complex camera motions, and varying lighting conditions.

MBA-SLAM: Motion Blur Aware Gaussian Splatting SLAM

Peng Wang*, Lingzhe Zhao*, Yin Zhang, Shiyu Zhao, Peidong Liu

Project ArXiv Code

• MBA-SLAM, the first deblurring SLAM that can accurately estimate the local camera motion trajectories within the exposure time and recovers the high quality 3D scene from blurry sequences.

BAD-Gaussians: Bundle Adjusted Deblur Gaussian Splatting

Lingzhe Zhao*, Peng Wang*, Peidong Liu

Project ArXiv Code

• BAD-Gaussians successfully deblurs severe motion-blurred images, synthesizes higher-quality novel views, and achieves real-time rendering, surpassing previous SOTA implicit deblurring rendering methods.