Han-Jia Ye

33

Papers

134

Total Citations

Papers (33)

What Makes Objects Similar: A Unified Multi-Metric Learning Approach

TabPFN Unleashed: A Scalable and Effective Solution to Tabular Classification Problems

Leveraging Cross-Modal Neighbor Representation for Improved CLIP Classification

MOS: Model Surgery for Pre-Trained Model-Based Class-Incremental Learning

Revisiting Nearest Neighbor for Tabular Data: A Deep Tabular Baseline Two Decades Later

Improving LLMs for Recommendation with Out-Of-Vocabulary Tokens

Understanding the Limits of Deep Tabular Methods with Temporal Shift

Addressing Imbalanced Domain-Incremental Learning through Dual-Balance Collaborative Experts

SIN: Selective and Interpretable Normalization for Long-Term Time Series Forecasting

Enhancing Class-Imbalanced Learning with Pre-Trained Guidance through Class-Conditional Knowledge Distillation

Tabular Insights, Visual Impacts: Transferring Expertise from Tables to Images

Multi-layer Rehearsal Feature Augmentation for Class-Incremental Learning

Few-Shot Learning via Embedding Adaptation With Set-to-Set Functions

Distilling Cross-Task Knowledge via Relationship Matching

Learning Placeholders for Open-Set Recognition

Forward Compatible Few-Shot Class-Incremental Learning

Identifying Ambiguous Similarity Conditions via Semantic Matching

Learning Debiased Representations via Conditional Attribute Interpolation

Procrustean Training for Imbalanced Deep Learning

FOSTER: Feature Boosting and Compression for Class-Incremental Learning

Dual Consolidation for Pre-Trained Model-Based Domain-Incremental Learning

Task-Agnostic Guided Feature Expansion for Class-Incremental Learning

External Knowledge Injection for CLIP-Based Class-Incremental Learning

Integrating Task-Specific and Universal Adapters for Pre-Trained Model-based Class-Incremental Learning

MIETT: Multi-Instance Encrypted Traffic Transformer for Encrypted Traffic Classification

Capability Instruction Tuning

Learning Only When It Matters: Cost

Expandable Subspace Ensemble for Pre-Trained Model-Based Class-Incremental Learning

A$^2$-Net: Learning Attribute-Aware Hash Codes for Large-Scale Fine-Grained Image Retrieval

Towards Enabling Meta-Learning from Target Models

Model Spider: Learning to Rank Pre-Trained Models Efficiently

Few-Shot Class-Incremental Learning via Training-Free Prototype Calibration

Rectify Heterogeneous Models with Semantic Mapping