scene understanding
Scene understanding involves the ability of AI systems to interpret and make sense of a visual environment. This includes recognizing objects, actions, and their relationships within a scene, often integrating spatial, semantic, and contextual information to create a coherent representation.
- C3Po: Cross-View Cross-Modality Correspondence by Pointmap Prediction
- CroPe: Cross-Modal Semantic Compensation Adaptation for All Adverse Scene Understanding
- FutureSightDrive: Thinking Visually with Spatio-Temporal CoT for Autonomous Driving
- Holistic Order Prediction in Natural Scenes
- HoloScene: Simulation‑Ready Interactive 3D Worlds from a Single Video
- IR3D-Bench: Evaluating Vision-Language Model Scene Understanding as Agentic Inverse Rendering
- Leader360V: A Large-scale, Real-world 360 Video Dataset for Multi-task Learning in Diverse Environment
- LiteReality: Graphic-Ready 3D Scene Reconstruction from RGB-D Scans
- MLLMs Need 3D-Aware Representation Supervision for Scene Understanding
- PhyBlock: A Progressive Benchmark for Physical Understanding and Planning via 3D Block Assembly
- Reinventing Multi-Agent Collaboration through Gaussian-Image Synergy in Diffusion Policies
- Robo2VLM: Improving Visual Question Answering using Large-Scale Robot Manipulation Data
- ST$^2$360D: Spatial-to-Temporal Consistency for Training-free 360 Monocular Depth Estimation
- Under the Shadow: Exploiting Opacity Variation for Fine-grained Shadow Detection
- UrbanIng-V2X: A Large-Scale Multi-Vehicle, Multi-Infrastructure Dataset Across Multiple Intersections for Cooperative Perception
- Visual Jenga: Discovering Object Dependencies via Counterfactual Inpainting
- mmWalk: Towards Multi-modal Multi-view Walking Assistance