Research team  ESAT - PSI, Centre for the Processing of Speech & Images

Research team description:

CENTER FOR PROCESSING SPEECH AND IMAGES (PSI) In this division research is concentrated on solving real world problems of image and speech processing: * The largest research domain is image processing (2D as well as 3D). Research is done on the analysis and processing of satellite images (teledetection), robot vision, visual inspection for industrial automation, medical image computing and analysis, image-guided surgery, 3D visualization and animation, processing of color images and image compression, and virtual and added reality. * Furhter there is also research on speech processing. The main focus is the development of a speaker independent recognizer for continuous speech with large vocabulary. This includes research on acoustic modeling, new techniques for preprocessing, search strategies and language modeling. In addition research is situated in the domain of corpus building, dialogue systems, speech coding and speech modification.
 

DIRAC: Detection and Identification of Rare Audio-visual Cues.
CLIF: Computational Linguistics in Flanders.
In vivo morphological, functional and molecular multimodality imaging of small animals. 
Context/scene-aware 3D modeling.
TELEX: Combining acoustic TEmplates en LEXical modeling.
Development and use of a 4D Geographic Information System to support the conservation of the Calakmu...
Feature Selection in Speech and Speaker Recognition
HI-MASQUERADE
Quantitative analysis of in vivo multimodal and multitemporal images: from animal models to novel me...
MRTN-CT-2006-035561 S2S Sound two Sense.
Spatial mutual information for non rigid registration.    
VASI: Vocabulary Acquisition from Speech and Images.   
MIDAS: MIssing DAta Solutions.
Model-based image computing for multi-X medical image data.
Development and valorization of novel methodologies for in vivo imaging in small animal models.
Model-based analysis of 3D+time cardiac medical images, using a hybrid segmentation and registration...

The active blind spot camera: hard real-time recognition of moving objects from a moving camera
Texture from multiple images
Object class recognition within a scalable probabalistic framework
Efficient Multiview and Multiclass Object Recognition
Quantification of bone architecture and vascularisation with Micro-CT.
Computational strategies for iconic shape modelling and model-based non-rigid image registration.
Sparse representations for automatic speech recognition
Multi-Class Multi-View Object Detection and Recognition
Automatische beeldgebaseerde localisatie en modellering: een geintegreerde benadering
Quantitative Analysis of Multi-Modal and Multi-Temporal In-Vivo Images of Animal Models
Generische 3D Modellen voor de Parametrisatie van het Menselijke Gezicht
Advanced vision processing and application-algorithm-architecture trade-offs in non-rigid, sparse mu...
Improved object recognition by combining visual and textual information
Acoustic Hypothesis Testing in an Exemplar-based Speech Recognize with Application to a Reading Tuto...
Inter-subject statistical shape modelling and analysis
Architectuur-gedreven 3D modellering van gebouwen
Segmentation, registration and visualization of the maxillofacial region based on cone-beam CT image...
Camera based Fall Detection System
Intra-subject deformation modeling
Modelling Partial Occlusion
Resolving the full 3D myocardial strain tensor in-vivo using ultrasound imaging
Computational strategies for the analysis of brain connectivity with DT-MRI
Robust speech recognition for large vocabulary using the missing data technique.
Graph-based statistical shape and appearance models for joint image segmentation and registration.
Image markers for Early Response Detection in Tumor Therapy
Large-scale image and video retrieval
Prediction of image quality in emission tomography
3D objectherkenning en objectcategorisatie
Multi-Modal Image Registration during Minimally Invasive Treatment
Speech recognition for media indexing and retrieval
Large Vocabulary Automatic Speech Recognition Using The Missing Data Paradigm
Multimodality non-rigid image registration for image-based radiotherapy.
Lifting 3D acquisition to a higher semantic level
Automatische herkenning van gesproken interviews
Visual Analysis of Human Actions
Accent Recognition for Forensic Applications
Segmentation of tree like structures by global optimisation.
Detection of abnormal behaviour in surveillance applications
Image based 3D reconstruction and its application
Relational Learning for Image Interpretation
Large-scale classification and retrieval of 3D shapes
-
3D building reconstruction using procedural modeling and high level information from object detector...
Cognitive image segmentation
Inverse procedural modeling of buildings
Analysis of multispectral brain images in the presence of pathology
Recognition of Buildings and Building Styles

Expertises of research team  ESAT - PSI, Centre for the Processing of Speech & Images :

Expertise SUETENS PAUL
- computer vision - medical image computing - 3-D computer graphics - image guided intervention - co...
Expertise WAMBACQ PATRICK
- stochastic signals analysis - pattern recognition - speech processing
Expertise VAN GOOL LUC
Consulting for industrial applications of computer vision. Courses on image processing and its appli...
Expertise VAN EYCKEN LUC
- Comparative study of compression techniques, including standards such as H2xx, JPEG, JBIG, MPEG - ...
Expertise VAN HAMME HUGO
- Speech processing. - Noise-robust speech recognition. - Speech recognition for learning applicatio...

Research team number: 50000519