Jinwoo Shin

14

Papers

437

Total Citations

Papers (14)

Representation Alignment for Generation: Training Diffusion Transformers Is Easier Than You Think

Discovering and Mitigating Visual Biases through Keyword Explanation

DreamFlow: High-quality text-to-3D generation by Approximating Probability Flow

Calibrated Multi-Preference Optimization for Aligning Diffusion Models

DiffusionGuard: A Robust Defense Against Malicious Diffusion-based Image Editing

Querying Easily Flip-flopped Samples for Deep Active Learning

Efficient Long Video Tokenization via Coordinate-based Patch Reconstruction

Adversarial Robustification via Text-to-Image Diffusion Models

Test-Time Adaptation with Binary Feedback

Data-Efficient Molecular Generation with Hierarchical Textual Inversion

Closest Neighbors are Harmful for Lightweight Masked Auto-encoders

Real-World Efficient Blind Motion Deblurring via Blur Pixel Discretization

Visual Representation Learning with Stochastic Frame Prediction

Controllable Human Image Generation with Personalized Multi-Garments