TPG-INR: Target Prior-Guided Implicit 3D CT Reconstruction for Enhanced Sparse-view Imaging

The introduction of Neighborhood Attention Filtering (NAF) represents a significant advancement in the field of Vision Foundation Models (VFMs), allowing for zero-shot feature upsampling without the need for retraining. This innovative method utilizes Cross-Scale Neighborhood Attention and Rotary Position Embeddings to adaptively learn spatial and content weights from high-resolution images, outperforming existing VFM-specific upsamplers across various tasks.

A new method called ReCoGS has been introduced for real-time recoloring of scenes using Gaussian Splatting, which is recognized for its efficiency in novel view synthesis and high-quality reconstructions. This user-friendly pipeline allows precise selection and recoloring of regions within pre-trained scenes, demonstrating real-time performance through an interactive tool. Code for the method is available online.

Recent advancements in image customization have been marked by the introduction of PositionIC, a framework designed to enhance fidelity and spatial control in multi-subject images. This development addresses the challenges posed by the lack of scalable, position-annotated datasets and the complexities of global attention mechanisms that entangle identity and layout. PositionIC incorporates BMPDS, an automatic data-synthesis pipeline, and a layout-aware diffusion framework with a novel visibility-aware attention mechanism.

AutoSAGE has been introduced as an input-aware CUDA scheduler designed to optimize sparse GNN aggregations, specifically SpMM and SDDMM, by dynamically selecting tiling and mapping strategies based on input characteristics. This innovation leverages lightweight estimates and on-device micro-probes, ensuring performance improvements while maintaining compatibility with vendor kernels.