Shiyu Chang

31

Papers

354

Total Citations

Papers (31)

Dilated Recurrent Neural Networks

NeurIPS 2017arXiv

Correcting Diffusion Generation through Resampling

Fictitious Synthetic Data Can Improve LLM Factuality via Prerequisite Learning

Decomposing Uncertainty for Large Language Models through Input Clarification Ensembling

D3: Deep Dual-Domain Based Fast Restoration of JPEG-Compressed Images

Studying Very Low Resolution Recognition Using Deep Networks

Image Super-Resolution via Dual-State Recurrent Networks

Additive Adversarial Learning for Unbiased Authentication

Adversarial Robustness: From Self-Supervised Pre-Training to Fine-Tuning

The Lottery Tickets Hypothesis for Supervised and Self-Supervised Pre-Training in Computer Vision Models

Quarantine: Sparsity Can Uncover the Trojan Attack Trigger for Free

Uncovering the Disentanglement Capability in Text-to-Image Diffusion Models

Robust Video Super-Resolution With Learned Temporal Dynamics

Jointly Attentive Spatial-Temporal Pooling Networks for Video-Based Person Re-Identification

AutoGAN: Neural Architecture Search for Generative Adversarial Networks

Harnessing the Spatial-Temporal Attention of Diffusion Models for High-Fidelity Text-to-Image Synthesis

Robust Mixture-of-Expert Training for Convolutional Neural Networks

VSP: Diagnosing the Dual Challenges of Perception and Reasoning in Spatial Planning Tasks for MLLMs

Sparse Cocktail: Every Sparse Pattern Every Sparse Ratio All At Once

Speech Self-Supervised Learning Using Diffusion Model Synthetic Data

Zeroth-Order Stochastic Variance Reduction for Nonconvex Optimization

A Game Theoretic Approach to Class-wise Selective Rationalization

Tight Certificates of Adversarial Robustness for Randomly Smoothed Classifiers

Training Stronger Baselines for Learning to Optimize

The Lottery Ticket Hypothesis for Pre-trained BERT Networks

Understanding Interlocking Dynamics of Cooperative Rationalization

TransGAN: Two Pure Transformers Can Make One Strong GAN, and That Can Scale Up

PARP: Prune, Adjust and Re-Prune for Self-Supervised Speech Recognition

Fairness Reprogramming

Selectivity Drives Productivity: Efficient Dataset Pruning for Enhanced Transfer Learning

AutoVC: Zero-Shot Voice Style Transfer with Only Autoencoder Loss