differentiable rendering
A rendering process that allows gradients to be computed with respect to scene parameters, enabling optimization of 3D scenes or parameters in AI models through backpropagation.
- DGH: Dynamic Gaussian Hair
- Flux4D: Flow-based Unsupervised 4D Reconstruction
- From Pixels to Views: Learning Angular-Aware and Physics-Consistent Representations for Light Field Microscopy
- GPSToken: Gaussian Parameterized Spatially-adaptive Tokenization for Image Representation and Generation
- Gaussian-Augmented Physics Simulation and System Identification with Complex Colliders
- LinPrim: Linear Primitives for Differentiable Volumetric Rendering
- Rendering-Aware Reinforcement Learning for Vector Graphics Generation
- TREND: Unsupervised 3D Representation Learning via Temporal Forecasting for LiDAR Perception