Fabien Baradel

I am a third year PhD Candidate at INSA Lyon - LIRIS.
I am working on Machine Learning and Computer Vision.
My supervisors are Christian Wolf and Julien Mille.
My PhD focuses on video understanding.

PhD thesis title:
"Deep Learning for Human Understanding:
poses, gestures, activities"

funded by the ANR/NSREC DeepVision project.

I received my Engineer's degree (MSc) from ENSAI with a major in Data Science. Previously I've been intern at Xerox Research Centre Europe.

Email  /  CV  /  Scholar  /  Github  /  LinkedIn  /  Twitter


My main researchs focus on video understanding (e.g. action recognition, human-object interaction). I am also interested into discovering causal concepts from video by leveraging the arrow of time.
Keywords: machine learning, deep learning, computer vision, domain adaptation, causality.


Object Level Visual Reasoning in Videos
Fabien Baradel, Natalia Neverova, Christian Wolf, Julien Mille Greg Mori
The IEEE European Conference in Computer Vision (ECCV), 2018
Project page / PDF / arXiv / video / bibtex / Code / Complementary Mask Data / Poster

A model capable of learning to reason about semantically meaningful spatio-temporal interactions in videos.


Human Activity Recognition with Pose-driven Attention to RGB
Fabien Baradel, Christian Wolf, Julien Mille
The British Machine Vision Conference (BMVC), 2018
PDF / bibtex / Poster

Human activity recogntion using skeleton data and RGB. We propose a network able to focus on relevant parts of the RGB stream given deep features extracted from the pose stream.


Glimpse Clouds: Human Activity Recognition from Unstructured Feature Points
Fabien Baradel, Christian Wolf, Julien Mille Graham Taylor
The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018
PDF / arXiv / project page / video / bibtex / CVPR Daily / Code / Poster

We propose a new method for human action recognition relying on RGB data only. A visual attention module is able to extract glimpses within each frame. Resulting local descriptors are soft-assigned to distributed workers which are finally classifying the video.


Human Action Recognition: Pose-based Attention draws focus to Hands
Fabien Baradel, Christian Wolf, Julien Mille
The IEEE International Conference on Computer Vision (ICCV), Workshop "Hands in Action", 2017
PDF / bibtex / Poster

A new spatio-temporal attention based mechanism for human action recognition able to automatically attend to most important human hands and detect the most discriminative moments in an action.


Discrepancy-based networks for unsupervised domain adaptation: a comparative study
Gabriela Csurka, Fabien Baradel, Boris Chidlovskii, Stephane Clinchant,
The IEEE International Conference on Computer Vision (ICCV), Workshop "Task-CV", 2017
PDF / bibtex

We introduce a new dataset for Domain Adaptation and show a comparaison between shallow and deep methods based on Maximum Mean Discrepancy.


Pose-conditioned Spatio-Temporal Attention for Human Action Recognition
Fabien Baradel, Christian Wolf, Julien Mille
arXiv preprint, 2017
arXiv / PDF / project page / video / bibtex

We introduce an attention-based mechanism around hands on RGB videos conditioned on features extracted from human 3D pose.


Conditional adaptation network for image classification
Fabien Baradel, Gabriela Csurka, Boris Chidlovskii,
US Patent App. 15/450,620 - Xerox Corp, 2017
PDF / bibtex

We introduce a new method based on Conditional Maximum Mean Discrepancy for domain adaptation on image classification.

Reviewer :ICML 2019, CVPR 2019, NIPS 2018, IJCV, TNNLS

Machine Learning
Science U - M2 Info - 28h30 (CM+TP) - 2018/2019
Github repo / Slides: 1 - 2 / Exercises: 1 - 2

Regression Modelling
Univ Lyon 1 - M2 Data Science - 12h (TP) - Fall 2017

Probability & Statistics
Univ Lyon 1 - L2 Info & Maths-Eco - 12h+8h (TP) - Fall 2017

EPITA - 1st year - 24h (CM+TD) - September 2017

Introduction to Deep Learning with Tensorflow
ENSAI - MSc Data Science - 6h - January 2017
Github repo / slides

Awesome webpage...