pose-prediction

[ICCV-2023] Official code for work "HumanMAC: Masked Motion Completion for Human Motion Prediction".

We evaluate our method on different datasets (including ShapeNet, CUB-200-2011, and Pascal3D+) and achieve state-of-the-art results, outperforming all the other supervised and unsupervised methods and 3D representations, all in terms of performance, accuracy, and training time.

[ICCV2023] Official PyTorch Implementation of "BeLFusion: Latent Diffusion for Behavior-Driven Human Motion Prediction". ICCV 2023

[ECCV 2020] Official PyTorch Implementation of "DLow: Diversifying Latent Flows for Diverse Human Motion Prediction". ECCV 2020.

Code and notebook for our paper "Assessing interaction recovery of predicted protein-ligand poses"

Comprehensive ensembling of protein-ligand structure and affinity prediction methods (CASP16)

A Set of Machine Learning Classifiers For RNA Containing Complexes

A measured benchmark of three docking scoring functions on 308 protein-ligand complexes, reporting pose accuracy and physical validity together, with the success rate recomputed after removing the crystal starting conformer and the ligand-centred search box.

Do structural priors help a co-folding model? PLIP constraints from crystals and DiffDock poses, fed to Boltz-2 and measured against an unconstrained baseline.

Fine-tuning OpenFold3 (4.0.0) for PDE10A protein–ligand pose prediction via distribution-aware PDB-scale data augmentation: +0.20 PL LDDT, −2.3 Å ligand RMSD on held-out.

This project predicts future human body poses from past pose sequences using the SMPL model encoded via VPoser. We model the task as a time-series prediction in VPoser latent space using a bidirectional GRU encoder-decoder.