LIFT

2026 Datasets

Dataset Analysis

Introduces LIFT, an image-to-video generation framework with layout-in-future and camera control, using on-policy self-distillation to learn from sparse last-frame layout annotations.

Provenance

Collected from papers.

Derived from paper: LIFT: Layout-In-Future Video Generation under Large Viewpoint Change via On-Policy Self-Distillation

Related papers

LIFT: Layout-In-Future Video Generation under Large Viewpoint Change via On-Policy Self-Distillation Exact equivariance, kept through training, buys zero-shot generalisation across the symmetry group How Should World Models Be Evaluated? A Decision-Making-Centric Position Latent Spatial Memory for Video World Models Lifting Embodied World Models for Planning and Control MoVerse: Real-Time Video World Modeling with Panoramic Gaussian Scaffold MosaicMem: Hybrid Spatial Memory for Controllable Video World Models MotionWAM: Towards Foundation World Action Models for Real-Time Humanoid Loco-Manipulation NarrativeWorldBench: A Frontier-Saturated Benchmark and a Latent World Model for Long-Horizon Co-Creative Audio Drama PLUME: Probabilistic Latent Unified World Modeling and Parameter Estimation for Multi-Finger Manipulation PanoWorld: A Generative Spatial World Model for Consistent Whole-House Panorama Synthesis Quantitative Video World Model Evaluation for Geometric-Consistency InfiniCube: Unbounded and Controllable Dynamic 3D Driving Scene Generation with World-Guided Video Models Image2Sim: Scaling Embodied Navigation via Generative Neural Simulator E3DGS: Unified Geometric-Photometric Equivariance for 3D Gaussian Splatting via Color-as-Geometry Embedding Lyra 2.0: Explorable Generative 3D Worlds WorldCycle: Self-Verifiable Reinforcement Learning for Long-Horizon Video World Models AtlasVLA: Persistent World-Ego State Modeling for Vision-Language-Action Models StateFlow: Building, Evolving, and Accessing 3D World States for Previsualization 4DAnyone: Create Anyone in 4D from a Casual Monocular Video WebWorld: The Browser as a World Model for Self-Improving Web Code Off-Manifold Refinement: Guiding Video Generators with a Frozen World Model InfluenceField: A Differentiable Field with Interventionally Identifiable Causal Structure for Multimodal World Modeling Compact Visuotactile World Models for Lifting: Prediction, Reward Alignment, and Force Constraints Genie Sim PanoWorld: An Infinite Indoor 3D World Generation Pipeline via Panoramic Scene Modeling and Simulation 2025. PanoWan: Lifting Diffusion Video Generation Models to 360 {\ deg } with Latitude/Longitude-aware Mechanisms RoboGaze: Evaluating Robot World Models via Structured Vision-Language Analysis SpaceCast-Bench: Evaluating Predictive Spatial Reasoning in Vision-Language Models LEGO: A Lifting-Free Approach for Exocentric-to-Egocentric Video Generation OuroWorld: Bringing Any 3D World Alive as Diverse, Endlessly Looping 3D Cinemagraphs

Resources

Paper