temporal consistency
Temporal consistency refers to the maintenance of coherence in outcomes across time in dynamic systems or processes. In AI, it is important in video processing and actions recognition to ensure logical progression of frames or events.
- Act Only When It Pays: Efficient Reinforcement Learning for LLM Reasoning via Selective Rollouts
- DenseDPO: Fine-Grained Temporal Preference Optimization for Video Diffusion Models
- FlowFeat: Pixel-Dense Embedding of Motion Profiles
- FrameShield: Adversarially Robust Video Anomaly Detection
- Hierarchical Fine-grained Preference Optimization for Physically Plausible Video Generation
- Improving Generative Behavior Cloning via Self-Guidance and Adaptive Chunking
- Learning Temporal 3D Semantic Scene Completion via Optical Flow Guidance
- LiveStar: Live Streaming Assistant for Real-World Online Video Understanding
- Neuro-Spectral Architectures for Causal Physics-Informed Networks
- One-Step Diffusion for Detail-Rich and Temporally Consistent Video Super-Resolution
- PPMStereo: Pick-and-Play Memory Construction for Consistent Dynamic Stereo Matching
- Real-Time Execution of Action Chunking Flow Policies
- ST$^2$360D: Spatial-to-Temporal Consistency for Training-free 360 Monocular Depth Estimation
- StateSpaceDiffuser: Bringing Long Context to Diffusion World Models
- TC-Light: Temporally Coherent Generative Rendering for Realistic World Transfer
- UniRelight: Learning Joint Decomposition and Synthesis for Video Relighting
- Unleashing the Potential of Multimodal LLMs for Zero-Shot Spatio-Temporal Video Grounding
- VividFace: A Robost and High-Fidelity Video Face Swapping Framework
- WorldWeaver: Generating Long-Horizon Video Worlds via Rich Perception