Anthony Cioppa
Ph.D. Student
I am currently working as a Ph. D. Student in the team of Prof. Marc Van Droogenbroeck. I am developing novel computer vision and deep learning based techniques for the problem of automatic video understanding and analysis in the context of sports videos.
Technical Skills
Education
University of Liège
Sep 2017 - Present
Ph.D.
Ph.D. Student in Computer Vision and Deep Learning at the University of Liège (ULiège), Belgium. Research centered on semantic segmentation and global video understanding in the case of sports videos. Financed by a four-year FRIA research grant (FRS - FNRS).
University of Liège
Sep 2015 - Jun 2017
Masters
Master’s degree in Eletrical Engineering specialized in Telecommunications and Signal Processing at ULiège.
University of Liège
Sep 2012 - Jun 2015
Bachelors
Bachelor’s degree in Civil Engineering at ULiège with Electricity and Physics as options.
Computer skills
Work Experience
Research Area
Sep 2017 - Present
Ph.D research
I am currently working as a Ph. D. candidate in the laboratory of Prof. Van Droogenbroeck. I am developing computer vision and deep learning-based techniques for the problem of automatic video understanding and analysis in the context of sports videos. Recently, I have been working on a novel method which allows to produce real-time segmentation masks of the soccer players in a video sequences using a deep learning network that adapts to the latest game conditions by continuously learning new features during the match (named ARTHuS, best paper award at CVPR 2019’s CVsports workshop). I coordinated this project with researchers from the University of Louvain. I also participated in the study of capsule networks and how they can be used to interpret the features captured by the layers (named HitNet, accepted at AAAI 2019’s workshop on network interpretability). Prior to these works, I worked on a bottom-up approach for classifying game events in soccer videos using low level features extracted with techniques that I developed. This paper won the best paper award at CVPR 2018’s workshop CVsports. At the moment, I am working on the integration of semantic information to improve the performances of background subtraction algorithms.
Teaching Assistant
Jan 2015 - Present
- Computer Vision: Some Theoretical Lectures, Practical Sessions and Responsible for students' projects. 2018 - Present
- Analog Electronics: Lab Assistant for the students. 2016 - 2017
- Analysis and Design of Electrical Measuring Systems: Lab Assistant for the students. 2015 - 2016
- Statistics: Responsible for students' projects. 2015 - 2016
- Probability: Responsible for students' projects. 2015
Awards an Grants
- Best Paper Award at the Computer Vision and Pattern Recognition Workshop (CVPRW) CVsports at CVPR 2019 in Long Beach, USA, sponsored by TrackMan. (First Author). 2019
- Best Paper Award at the Computer Vision and Pattern Recognition Workshop (CVPRW) CVsports at CVPR 2018 in Salt Lake City, USA, sponsored by Second Spectrum. (First Author). 2018
- FRIA Research Grant. Research grant granted by the FRS - FNRS aiming at the realization of a doctoral thesis in four years. 2018
- Best Master’s Thesis Award. Granted by the AIM (Association des Ingénieurs de Montefiore) for my Master‘s Thesis. 2017
- Prize for the best Master’s Thesis. Granted by the AILg (Association des Ingénieurs diplômés de l’Université de Liège) for the same Master’s Thesis. 2017
Publications
- A. Cioppa, A. Deliège, M. Istasse, C. De Vleeschouwer, and M. Van Droogenbroeck. “ARTHuS: Adaptive Real-Time Human Segmentation in Sports through Online Distillation” Accepted for IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) 2019.
- A. Deliège, A. Cioppa, and M. Van Droogenbroeck. “An Effective Hit-or-Miss Layer Favoring Feature Interpretation as Learned Prototypes Deformations.” In AAAI Conference on Artificial Intelligence, Workshop on Network Interpretability for Deep Learning, 2019.
- A. Cioppa, A. Deliège, and M. Van Droogenbroeck. “A bottom-up approach based on semantics for the interpretation of the main camera stream in soccer games.” In IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) 2018. (Best paper Award).
- A. Deliège, A. Cioppa, and M. Van Droogenbroeck. "HitNet: a neural network with capsules embedded in a Hit-or-Miss layer, extended with hybrid data augmentation and ghost capsules." arXiv, 2018
- M. Van Droogenbroeck and A. Cioppa and A. Deliège. “Image classification using neural networks.” European Patent Office. Patent pending.
Portfolio
ARTHuS: Pipeline
HitNet: Architecture
Contact
University of Liège
B28 Montefiore
Office: 2.99
10 Allée de la découverte
4000 Liège
Belgium
Tel. +32 4 366 26 41