ECCV 2026 accepted papers on Probabilistic and scientific modeling
38 of the 2,834 papers accepted at ECCV 2026 (1.3%) are on Probabilistic and scientific modeling, compared with 0.9% at ECCV 2024. They fall into 6 topics.
See ECCV 2026 in the interactive map
Working on something in this area? Describe your idea in scime atlas to see the closest papers by meaning and how crowded the spot has become.
Topics and papers
Variational and generative sampling · 8 papers
- SimFlow: Simplified and End-to-End Training of Latent Normalizing Flows (8 citations)
- Entropy-Controlled Flow Matching (2 citations)
- Posterior Augmented Flow Matching (1 citation)
- ReSWD: ReSTIR‘d, not shaken. Combining Reservoir Sampling and Sliced Wasserstein Distance for Variance Reduction. (1 citation)
- A Scalable Vector Graphics Latent Space
- GeoCFM: Positive-Only Conditional Flow Matching for Mineral Occurrence Sampling
- Parsimonious Flow Matching for Efficient Image Generation
- VNC: A Scale-Space Foundation for Learnable 3D Surface Evolution
Dynamical systems · 8 papers
- RainODE: Continuous-Time Precipitation Forecasting with Latent Neural ODEs (2 citations)
- Adaptive Neural Dynamics for Robust Geometric LiDAR-Inertial State Estimation on UAVs
- NeuIDO: Neural Intrinsic Dynamics Operator for Physics-Informed 4D World Models
- Physics-Grounded Disentangled Flow Modeling for Brain Disease Progression Trajectory
- PMGC-SimVP: Parametric Multi-scale Gated Convolution for Global Ionospheric TEC Prediction
- Progression as Latent Drift: Generative Forecasting of Slow-Evolving Pathologies
- CLARITY: Medical World Model for Guiding Treatment Decisions by Simulating Context-Aware Disease Trajectories in Latent Space
- NeuralDMD: Interpretable Untrained Neural Network for Imaging from Sparse and Noisy Observations
Differential equations and PDEs · 7 papers
- ELHINN: Unifying Dense Crowd Simulation Across Scales via Eulerian–Lagrangian Hydrodynamics
- Markov-Renewal Single-Photon LiDAR Simulator
- Physics-Guided Deep Learning for Linear Mueller Matrix Acquisition
- Puppet-CNN: Continuous Parameter Dynamics for Input-Adaptive Convolutional Networks
- Two-Parameter Flow Map Learning for Continuous-Time Diffeomorphic Image Registration
- FUSE: Filter-Free Unified Spatiotemporal Estimation of SpO2 via Wave-Transport Modeling
- Learning Physics-based Forward Model Corrections in Unrolled Networks for Diffuser-based Imaging
Molecules and proteins · 5 papers
- Social-Mamba: Socially-Aware Trajectory Forecasting with State-Space Models (2 citations)
- CCFM: Collision-Constrained Flow Matching for Safety-Critical Scenario Generation
- PersonaDrive: Controllable Trajectory Prediction with Multi-Dimensional Driving Personas
- Towards Metric-Agnostic Trajectory Forecasting
- Unveiling Transferability in Trajectory Prediction via Latent Scene Embeddings
Causal discovery · 5 papers
- CausalVAE as a Plug-in for World Models: Towards Reliable Counterfactual Dynamics (3 citations)
- SPARC: Scalable Path-Specific Counterfactual Fairness via Causal Conditional Independence (1 citation)
- Beyond the Black Box: Identifiable Interpretation and Control in Generative Models via Causal Minimality
- Causal Intervention in Concept Bottleneck Models
- Decoding Multimodal Causality: End-to-End Multimodal Mediation Pathways Inference
Uncertainty and Gaussian processes · 5 papers
- Analytic Bayesian Uncertainty for LiDAR Segmentation: A Single-pass Generative Approach
- Bayesian Uncertainty Attribution-Guided Fine-Tuning for Open-Set Action Recognition
- SPARC: Single-Pass Scaling for Motion Forecasting with Conformal Bayesian Last Layers
- Structured SIR: Efficient and Expressive Importance-Weighted Inference for High-Dimensional Image Registration
- FuDU: A Fuzzy Dual-dimension Uncertainty Framework for Streaming Active Learning in Industrial Defect Detection
