Laplacian Pyramid, Taylor approximation principle, Dehazing, Interpretability
Code Generation, Multi-stage, large language models, Prompt Technique
Code Summarization, Local API Dependency Graph, AST, GCN
4K-HAZE dataset, CS-Mixer, GU-Net, a single GPU, real-time
low-light image enhancement, frequency-spatial network, fast fourier transform, random masking, fourier loss, adaptive gated fusion mechanism.
Knowledge Distillation, Image Restoration Security, Singular Value Perturbation, Undistillationn Defense
image enhancement, contrastive learning, multi-scale, channel attention
label distribution learning, Implicit Neural Representation, Gaussian Prior Constraints
Pedestrian trajectory prediction, Atmospheric Scattering Model, Graph Neural Networks, state space model, Mamba
