Keep on to blur preview images; turn off to show them clearly
Paper: https://t.co/MXLLZ0jzG7 Project: https://t.co/xxWzCSorHS Code: https://t.co/50yINVq1m1
Founder and CEO of https://t.co/5MjtfpwEU3 | Your guide to radiance fields | Host of the podcast @ViewDependent | FTP: 279 | discord: https://t.co/lrl64WGvlD
SkelSplat: Robust Multi-view 3D Human Pose Estimation with Differentiable Gaussian Rendering Contributions: 1. We propose SkelSplat, a novel framework for multi-view 3D human pose estimation, leveraging differentiable Gaussian rendering for view fusion. 2. We adapt Gaussian Splatting, primarily used for dense scene modeling, to skeleton-based 3D pose estimation. 3. We modify the original Gaussian Splatting rendering function to encode human joints using a one-hot representation, enabling pose-specific optimization. 4. We demonstrate that SkelSplat achieves accurate 3D pose estimation under challenging occlusions and varying camera setups, without requiring retraining or fine-tuning.
Paper: https://t.co/MXLLZ0jzG7 Project: https://t.co/xxWzCSorHS Code: https://t.co/50yINVq1m1
RT @DSkaale: BIG NEWS: Introducing SDF-Splats! 🚨 I just solved one of the biggest problems in 3D Gaussian Splatting: GAPS & HOLES! 🕳️❌ My…
Founder and CEO of https://t.co/5MjtfpwEU3 | Your guide to radiance fields | Host of the podcast @ViewDependent | FTP: 279 | discord: https://t.co/lrl64WGvlD
RT @janusch_patas: First Corporate Sponsor: Core11 GmbH Supports LichtFeld Studio! Today marks an exciting milestone for LichtFeld Studio:…
Founder and CEO of https://t.co/5MjtfpwEU3 | Your guide to radiance fields | Host of the podcast @ViewDependent | FTP: 279 | discord: https://t.co/lrl64WGvlD
ConeGS: Error-Guided Densification Using Pixel Cones for Improved Reconstruction with Fewer Primitives TL;DR: "ConeGS replaces cloning-based densification with a novel method that generates pixel-cone-sized primitives in regions of high image-space error. By improving placement and removing reliance on existing scene structure—thanks to a flexible iNGP-based exploration—it achieves higher reconstruction quality than baselines using the same number of primitives." Contributions: • A densification strategy that places new Gaussians in regions of high photometric error in image space, guided by depth estimates from an iNGP-based geometric proxy. • An approach that determines the size of new Gaussians from the viewing cones of the pixels from which they are generated. • An improved opacity penalty that promptly removes low-opacity Gaussians, combined with a budgeting strategy that balances scene complexity and primitive count.
Paper: https://t.co/bXah0hRKW5 Project: https://t.co/jngGBkvwAv
YoNoSplat: You Only Need One Model for Feedforward 3D Gaussian Splatting TL;DR: YoNoSplat reconstructs 3D Gaussian splats directly from unposed and uncalibrated images, while flexibly leveraging ground-truth camera poses or intrinsics when available. Contributions: • We introduce YoNoSplat, the first feedforward model to achieve state-of-the-art performance in both pose-free and pose-dependent settings for an arbitrary number of views. • We identify the entanglement of pose and geometry learning as a key challenge and propose a mix-forcing training strategy that effectively mitigates training instability and exposure bias. • We resolve the scale ambiguity problem through an intrinsic-prediction-and-conditioning pipeline and a pairwise distance normalization scheme, enabling reconstruction from uncalibrated images.
Paper: https://t.co/48KjtPffc0 Project: https://t.co/9GTx4Lziic