reconstruction fidelity
A measure of how accurately a model can reconstruct original data from its compressed form or latent representation. It is crucial in applications such as image compression and denoising, where maintaining original data quality is essential.
- DC4GS: Directional Consistency-Driven Adaptive Density Control for 3D Gaussian Splatting
- I2-NeRF: Learning Neural Radiance Fields Under Physically-Grounded Media Interactions
- Latent Harmony: Synergistic Unified UHD Image Restoration via Latent Space Regularization and Controllable Refinement
- LinPrim: Linear Primitives for Differentiable Volumetric Rendering
- MoRE-Brain: Routed Mixture of Experts for Interpretable and Generalizable Cross-Subject fMRI Visual Decoding
- OmniFC: Rethinking Federated Clustering via Lossless and Secure Distance Reconstruction
- One-Step Diffusion-Based Image Compression with Semantic Distillation
- Orientation-anchored Hyper-Gaussian for 4D Reconstruction from Casual Videos
- PathVQ: Reforming Computational Pathology Foundation Model for Whole Slide Image Analysis via Vector Quantization
- ReCon-GS: Continuum-Preserved Guassian Streaming for Fast and Compact Reconstruction of Dynamic Scenes
- Sparc3D: Sparse Representation and Construction for High-Resolution 3D Shapes Modeling
- VaporTok: RL-Driven Adaptive Video Tokenizer with Prior & Task Awareness