An Innosuisse project allowed Idiap to set up AI tools able to help experts to study the manufacturability of new aluminum elements requested by their customers.
Perception & Activity Understanding
The analysis of human activities from multimodal data is useful for surveillance, behavior analysis, human–robot interfaces, and multimedia content analysis. This includes investigating the fundamental tasks of scene analysis such as detection, segmentation and tracking of people, their representation, and the characterization of their condition, as well as the modeling of sequential data and their interpretation in the form of gestures, activities, behavior, or social relationships, through the design of sound algorithms which exploit and extend models and methods of computer vision, machine learning, and multimodal data-fusion. Surveillance, traffic analysis, analysis of behavior, human-robot interfaces, and multimedia content analysis are the main application domains.
Group News
Idiap Research Institute and the School of Engineering at EPFL invite applications for the directorship of Idiap. The successful candidate will also hold a faculty position as full professor at EPFL School of Engineering.
Suraj Srinivas and Weipeng He received the 2021 PhD Thesis Distinction in Electrical Engineering by the EPFL.
Artificial intelligence and big data pioneers Octopeek (France) and Idiap announce a partnership. A member of Octopeek’s scientific staff will spend four years at the very heart of Idiap, culminating in a PhD—a unique opportunity to use video data to develop research on multimodal learning, and to spur innovation.
Human-robot interactions are often lacking fluidity, especially outside of the lab. Today, researchers from Idiap are publishing in open access the algorithms which allowed a robot to be used in real conditions in a shopping mall in Finland in the framework of the European project Mummer.
Group Job Openings
Our group is regularly posting job openings ranging from internships to researcher positions. To check the opportunities currently available or to submit a speculative applications use the link below.Our group is regularly posting job openings ranging from internships to researcher positions. To check the opportunities currently available or to submit a speculative applications use the link below.
Current Group Members
The group is led Jean-Marc Odobez.

ODOBEZ, Jean-Marc
(Senior Research Scientist with Academic Title)
- website

VILLAMIZAR, Michael (Alejandro)
(Research Associate)
- website

GUPTA, Anshul
(Research Assistant)
- website

VUILLECARD, Pierre
(Research Assistant)
- website

TAFASCA, Samy
(Research Assistant)
- website

CHUTISILP, Naravich
(Research Intern)

FARKHONDEH, Arya
(Research Intern)
- website

DE CAMPOS, Ruben
(AI Master Student)
- website
Alumni
- ALI, Abid
- AMINIAN, Bozorgmehr (Nima)
- BA, Silèye
- CAN, Gulcan
- CAO, Yuanzhouhan
- CHAVANE, Cécile
- CHEN, Cheng
- CHEN, Yiqiang
- CROVETTO, Gianna Larissa
- DE OLIVEIRA PRESTES, Lukas
- DESPRÈS, Nicolas
- DUVAL, Matthieu
- EMONET, Rémi
- FUNES MORA, Kenneth Alberto
- GAY, Paul
- GERVAISE, Lara (Audrey, Yasmina)
- GEVERS, Louis
- HE, Weipeng
- HEILI, Alexandre
- HU, Rui
- KHALIDOV, Vasil
- LE, Nam
- LEFÈVRE, Stéphanie
- LIU, Gang
- LOPEZ-MENDEZ, Adolfo
- MAHMOUDIAN, Navid
- MOSTAANI, Zohreh
- RACCA, Mattia
- RICCI, Elisa
- ROMAN RANGEL, Edgar Francisco
- SCHEFFLER, Carl
- SHEIKHI, Samira
- SIEGFRIED, Rémy
- SOUSA EWERTON, Marco (Antonio)
- STEL, Lucas
- TAVENARD, Romain
- VARADARAJAN, Jagannadan
- WU, Di
- YAO, Jian
- YU, Yu
Active Research Grants
Past Research Grants
- ASLEEP - Adapting the Static Luggage dEtEction module for the ProtectRail demonstrator
- EUMSSI - EUMSSI - Event Understanding through Multimodal Social Stream Interpretation
- G3E - G3E: Geometric Generative Gaze Estimation model
- GAZESENSESCREEN - GazeSense Scren
- HAI-2010 - Human activity and interactivity modeling
- HEAP - Human-Guided Learning and Benchmarking of Robotic Heap Sorting
- HUMAVIPS - Humanoids with auditory and visual abilities in populated spaces
- IMPACT - Image Spam Classification
- LIFE - PdG-fatigue
- MUMMER - MultiModal Mall Entertainment Robot
- P3 - P3: Press Pressure Prediction
- PROMOVAR - Probabilistic Motifs for Video Action Recognition
- REGENN - Robust Eye-Gaze Estimation Deep Neural Network
- ROSALIS - Robot skills acquisition through active learning and social interaction strategies
- SODA - Person Recognition in debate and broadcast news
- TRACOME - Robust face tracking, feature extraction and multimodal fusion for audio-visual speech recognition
- UNICITY - 3D scene understanding through machine learning to secure entrance zones
- VANAHEIM - Video/Audio Networked surveillance system enhAncement through Human-cEntered adaptIve Monitoring
- VIDEOPROTECTOR - Morphean VideoProtector
- VIEW-2 - Visibility Improvement for Events Webcasting