Jiajun Liang

15

Papers

143

Total Citations

Papers (15)

Improving Video Generation with Human Feedback

A Simple Baseline for Efficient Hand Mesh Reconstruction

MegActor-Sigma: Unlocking Flexible Mixed-Modal Control in Portrait Animation with Diffusion Transformer

Decoupled Knowledge Distillation

Boosting Semi-Supervised Learning by Exploiting All Unlabeled Data

Implicit Identity Leakage: The Stumbling Block to Improving Deepfake Detection Generalization

DOT: A Distillation-Oriented Trainer

Cumulative Spatial Knowledge Distillation for Vision Transformers

Efficient One Pass Self-Distillation with Zipf’s Label Smoothing

Explaining Deepfake Detection by Analysing Image Matching

Discriminability-Transferability Trade-Off: An Information-Theoretic Perspective

Joint Token Pruning and Squeezing Towards More Aggressive Compression of Vision Transformers

EAST: An Efficient and Accurate Scene Text Detector

Dynamic MLP for Fine-Grained Image Classification by Leveraging Geographical and Temporal Information

Sharp Impossibility Results for Hyper-graph Testing