Experience the future of computer vision with our advanced 3D Gaussian Splatting pipeline that converts flat 2D images
into immersive 3D representations.
Developed at KAUST, this revolutionary project combines SAM segmentation, Depth-Anything-V2, and COLMAP
to bridge the gap between traditional photography and three-dimensional visualization.
Our innovative graduation project at King Abdullah University of Science and Technology (KAUST) represents a breakthrough in 3D reconstruction and computer vision. We've developed a comprehensive pipeline that combines 3D Gaussian Splatting with advanced segmentation and depth estimation techniques to transform 2D images into detailed 3D models with remarkable accuracy.
Combines SAM segmentation, Depth-Anything-V2, and 3D Gaussian Splatting
Efficient 3D Gaussian Splatting with accelerated rendering
Precise depth estimation using Depth-Anything-V2 and COLMAP
Cutting-edge research combining multiple state-of-the-art technologies
See how our 3D Gaussian Splatting pipeline converts flat images into detailed 3D models
