Idiap combines its multi-diciplinary expertise to advance the understanding of human perceptual and cognitive systems, engaging in research on multiple aspects of human-computer interaction with computational artefacts such as natural language understanding and translation, document and text processing, vision and scene analysis, multimodal interaction, computational cognitive systems, and methods for automatically training such systems (see our research efforts in machine learning).
Speech and audio processing, computer vision, handwriting recognition, document processing, computational cognitive sciences, robotics, natural language processing, machine translation
Social Signal Processing is the domain aimed at automatic understanding of social interactions through analysis of nonverbal behavior
Web and mobile social media, social interaction sensing, social signal processing, verbal and nonverbal communication analysis, mobile phone sensing, computational social science
Information processing by computers must be accompanied by the capacity to present results to users in an efficient and usable way, using human-computer interfaces. In the case of interactive systems, these interfaces must also allow users to enter information in a simple and reliable way, and in the most advanced cases to acquire information from users in a non-disruptive ways. Current research directions at Idiap focus on multimedia information systems, user interfaces, and the evaluation of interactive systems, explained below in more detail.
Multimedia information systems, user interfaces, contextualization, personalization, system evaluation, mobile HCI using Big Data, data driven services
Besides multimodal (audio, speech, vision, social media) data access, Idiap is also very active in protecting data access, going beyond passwords and PIN codes. Biometric person recognition refers to the process of automatically recognizing a person using distinguishing behavioral patterns or physiological traits (like face, voice, fingerprints, etc). Current research directions include advanced face verification, speaker verification, joint face-speaker verification, new audio-visual features, and evaluation metrics. More recently, new research efforts were also deployed towards direct (spoofing) attacks to trusted biometric systems.
Speaker recognition, face recognition, multimodal biometric fusion, mobile biometry, spoofing and anti-Spoofing
Research in machine learning aims at developing computer programs able to learn from examples. Instead of relying on a careful tuning of parameters by human experts, machine learning techniques use statistical methods to directly estimate the optimal setting, which can hence have a complexity beyond what is achievable by a human experts.
Statistical and neural network based ML, computational efficiency, online learning, multi-sensor processing, very large datasets