panosantoniadis / ntua-abaw2021

[ICCVW 2021] Official implementation: An audiovisual and contextual approach for categorical and continuous emotion recognition in-the-wild

Hi, my name is Panagiotis Antoniadis 👋

I am an ELLIS PhD student in Machine Learning at the University of Copenhagen, supervised by Ole Winther. My research focuses on applying deep learning methods to biological applications. Previously, I worked at DeepLab, where I developed generative models for transcriptomics and deep neural networks for EEG-based brain-computer interfaces. I hold a BSc and MSc degree in Electrical and Computer Engineering from NTUA, where I completed my thesis on visual emotion recognition under the supervision of Petros Maragos.

ntua-abaw2021's Issues

Use of OpenPose for extracting Body and Face data

Hi Authors
I want to replicate your results for another research project. However, I cannot figure out how you used the openpose model for face and body extraction on the dataset. It would be helpful if you could provide me with some guidance on it.
Currently, I am referring to the https://github.com/CMU-Perceptual-Computing-Lab/openpose/blob/master/doc/installation/0_index.md#windows-portable-demo installing guide, but since I am running it on a remote server, the GUI interface could not be initialized.

Pretrained Models for face and context Backbones

Hi Authors
Could you provide the pre-trained models for face and context to be used as ConvNet Backbones.