An electron-density point-cloud framework for robust protein–ligand interaction prediction

Yujian Liu^*, Yutong Wang^*, Qingquan Wang^*, Meitang Peng, Yuan Chen, Yuechuan Lin, Dongxu Shen, Xiaoli Liu, Shidang Xu^†, Bin Liu

Predicts protein–ligand binding affinity from electron-density point clouds, remaining robust when atomic coordinates are noisy or come from predicted structures.

Details Paper Code

AnyAvatar: High-Fidelity Gaussian Head Avatars under Uncalibrated Camera Settings

Yujian Liu^*, Dongxu Shen^*, Haoran Li, Yuting Liu, Chuang Chen, Xinyi Jiang, Zhupeng Jiang, Peng Cao, Shidang Xu, Xiaoli Liu^†

Reconstructs animatable 3D Gaussian head avatars from uncalibrated multi-view images by jointly refining camera poses, FLAME geometry, and Gaussian appearance.

Details Project Code

MPFusion-MIL: Morphology-Guided Fusion with Precise Cross-Scale Interaction for Whole Slide Image Analysis

Yujian Liu, Ruoxuan Wu, Yuechuan Lin, Xinjie Shen, Yutong Wang, Haiyu Zhou, Shipu Xu, Shaoai Cai, Lingyu Liang, Shidang Xu^†

A multi-magnification MIL model that keeps spatial correspondence across scales and injects localized high-resolution evidence into aligned coarse regions.

Details

PyraE2E: Enhancing End-to-End WSI Analysis via Cross-Scale Super-Resolution

Yuechuan Lin^*, Yujian Liu^*, Weipeng Zhang, Yanyu Fan, Zikang Wang, Dongxu Shen, Liqin Fei, Xiaoli Liu, Shidang Xu^†

An end-to-end WSI framework that turns the slide resolution pyramid into cross-scale super-resolution supervision, so the encoder and aggregator can be trained together.

Details

SyncAnimation: A Real-Time End-to-End Framework for Audio-Driven Human Pose and Talking Head Animation

Yujian Liu^*, Shidang Xu^*, Jing Guo, Dingbin Wang, Zairan Wang, Xianfeng Tan, Xiaoli Liu^†

A real-time NeRF talking avatar that jointly synthesizes audio-synchronized head pose, facial expression, and lip motion in one pipeline.

Details Poster Project Code arXiv

Minimal High-Resolution Patches Are Sufficient for Whole Slide Image Representation via Cascaded Dual-Scale Reconstruction

Yujian Liu^*, Yuechuan Lin^*, Dongxu Shen^*, Haoran Li, Yutong Wang, Xiaoli Liu, Shidang Xu^†

Shows that a small set of informative high-resolution patches, selected and reconstructed through cascaded dual-scale learning, is sufficient for robust WSI representation.

Details Poster arXiv

MoGaFace: Momentum-Guided and Texture-Aware Gaussian Avatars for Consistent Facial Geometry

Yujian Liu, Linlang Cao, Chuang Chen, Fanyu Geng, Dongxu Shen, Peng Cao, Shidang Xu^†, Xiaoli Liu^†

Improves 3D Gaussian head avatars by jointly correcting facial geometry and recovering texture during rendering, instead of relying on a frozen tracked mesh.

Details Poster Project arXiv

Machine Learning-Enhanced Nanoparticle Design for Precision Cancer Drug Delivery

Qingquan Wang, Yujian Liu, Chenchen Li, Bin Xu, Shidang Xu^†, Bin Liu^†

Reviews how machine learning can guide nanoparticle synthesis and formulation, and help model nano–bio interactions along the cancer drug-delivery pipeline, from circulation and tumor extravasation to penetration and cellular uptake.

Paper

Discovery of activable oncolytic ionizable lipid nanoparticles for selective cancer therapy

Hou-Bing Zhang^*, Yujian Liu^*, Weide Xu^*, et al.

Uses machine-learning-accelerated screening to discover activable oncolytic ionizable lipid nanoparticles for selective cancer therapy.

DXTalker: Factorizing Speech-Driven 3D Facial Animation via Articulatory Prototypes and Personalized Dynamics

Yujian Liu, Shidang Xu, Xiaoli Liu, et al.

Factorizes speech-driven 3D facial animation into shared articulatory prototypes and identity-specific dynamics for accurate lip sync and expressive motion.