Official implementation of the paper "Estimation of continuous valence and arousal levels from faces in naturalistic conditions", Antoine Toisoul, Jean Kossaifi, Adrian Bulat, Georgios Tzimiropoulos and Maja Pantic, Nature Machine Intelligence, 2021

Generates multi-instrument symbolic music (MIDI), based on user-provided emotions from valence-arousal plane.

[NeurIPS 2022] The official repository of Expression Learning with Identity Matching for Facial Expression Recognition

This project is a semi-supervised approach to detect emotions on faces in-the-wild using GAN

A Unimodal Valence-Arousal Driven Contrastive Learning Framework for Multimodal Multi-Label Emotion Recognition (ACM MM 2024 oral)

[ECCV2022] The official repository of Emotion-aware Multi-view Contrastive Learning for Facial Emotion Recognition

Emotion recognition with Keras library. Uses AffectNet dataset and valence-arousal labels. Implements CNN architecture with regression

This Repo consist code for transfer learning for facial emotion detection via valence and arousal levels. We used pretrained weights from VGG-16 net and apply on that features deep neural network and lstm model in pytorch. We tested our model on Aff-wild net dataset.

[AAAI2021] A repository of Contrastive Adversarial Learning for Person-independent FER

Project to find one of 9 pre-trained emotions on given photo, video or webcam-stream

Code for "It’s LeVAsa not LevioSA! Latent Encodings for Valence-Arousal Structure Alignment" [CoDS-CoMAD'21]

[CHI2021] Hidden emotion detection using multi-modal signals

hsemotion의 ONNX 구현입니다.

An Emotional Game for Social Expression Learning

Search for a dependency between "annemo" annotations of valence and arousal from RECOLA and Action Units from OpenFace using ARD regression and ARIMA.

A hybrid CNN-RNN model trained on the DEAM dataset for music emotion recognition using the valence-arousal model. Useful for tagging emotional content of music in media or recommendation systems

Experiment to better understand how to use emotion analysis to understand a person's emotional reaction to an audiovisual work

Experiments with convolutional neural networks for emotion recognition

Multitask Affect Recognition with CNNs — Deep Learning Assignment 1 (CS452). Implements ResNet-18 and EfficientNet-B0 for facial expression classification (8 classes) and valence/arousal regression, with training graphs, metrics (Accuracy, F1, Kappa, Alpha, AUC, RMSE, CORR, SAGR, CCC), and report.

A Hybrid CNN-ViT Approach for Quantitatively Decoding & Modeling of Pet Sentiments / Instincts from Visual Data