Deyu Meng

Unveiling the Mechanism of Continuous Representation Full-Waveform Inversion: A Wave Based Neural Tangent Kernel Framework

Separable Neural Networks: Approximation Theory, NTK Regime, and Preconditioned Gradient Descent

Tucker-FNO: Tensor Tucker-Fourier Neural Operator and its Universal Approximation Theory

Sharper Analysis of Single-Loop Methods for Bilevel Optimization

NeuMatC: A General Neural Framework For Fast Parametric Matrix Operation

Efficient Conformal Prediction via Conformalized Meta-Learning from Noisy Labels

Exploring Imbalanced Annotations for Effective In-Context Learning

Polyline Path Masked Attention for Vision Transformer

Online Functional Tensor Decomposition via Continual Learning for Streaming Data Completion

SD-LoRA: Scalable Decoupled Low-Rank Adaptation for Class Incremental Learning

Improving Memory Efficiency for Training KANs via Meta Learning

Spatial-Mamba: Effective Visual State Space Models via Structure-Aware State Fusion

Semi-Supervised Regression with Heteroscedastic Pseudo-Labels

Feature Alignment with Equivariant Convolutions for Burst Image Super-Resolution

HAIR: Hypernetworks-based All-in-One Image Restoration

Diversity-Enhanced and Classification-Aware Prompt Learning for Few-Shot Learning via Stable Diffusion

Singular Value Fine-tuning for Few-Shot Class-Incremental Learning