Causal-Ex: Causal Graph-based Micro and Macro Expression Spotting

[ECCV] HYDRA: A Hyper Agent for Dynamic Compositional Visual Reasoning

Explain Before You Answer: A Survey on Compositional Visual Reasoning

[ICCV25] DWIM: Towards Tool-aware Visual Reasoning via Discrepancy-aware Workflow Generation & Instruct-Masking Tuning

Source code for the paper 'Beyond Efficiency: Molecular Data Pruning for Enhanced Generalization' (NeurIPS 2024).

Pruning By Explaining Revisited: Optimizing Attribution Methods to Prune CNNs and Transformers, Paper accepted at eXCV workshop of ECCV 2024

This repository gathers the SchNet4AIM code along with some instructions and readme files.

[WACV 2024] FG-Net: Facial Action Unit Detection with Generalizable Pyramidal Features

Face recognition with deep neural networks.

CNN LSTM architecture implemented in Pytorch for Video Classification

Official PyTorch implementation of the paper: UniGaze: Towards Universal Gaze Estimation via Large-scale Pre-Training.

Official repository for LTX-Video

A Lightweight Face Recognition and Facial Attribute Analysis (Age, Gender, Emotion and Race) Library for Python

cG-SchNet - a conditional generative neural network for 3d molecular structures

G-SchNet extension for SchNetPack

Tensorflow implementation of MolGAN: An implicit generative model for small molecular graphs

[CVPR 2023] Official repository of paper titled "Fine-tuned CLIP models are efficient video learners".

Official Pytorch implementation of the paper "MotionCLIP: Exposing Human Motion Generation to CLIP Space"

Code for the paper: Counterfactual Discriminative Micro-Expression Recognition

Post hoc Adversarial Stickers Against Expression Leakage (ICASSP25)

🤗 Diffusers: State-of-the-art diffusion models for image, video, and audio generation in PyTorch.

Machine Learning Applications in Wireless Communications - Project work

Generating optical flow frame by using TVL1 algorithm in organized way is not so difficult. But for a newbie it is toilsome to find this easier task. TVL1 is most popular way to generate optical fl…

Boosting Synthetic Data Generation with Effective Nonlinear Causal Discovery

Created a classifier that can identify and classify 41 types of facial attributes, using an ImageNet pre-trained ResNext50 model provided by PyTorch and CelebA dataset. Multiple data augmentation t…

Benchmark study of quality and faithfulness of counterfactual image generation

The codes and models in 'Gaze Estimation using Transformer, ICPR2022'.