A5059620912- Advanced Image Processing Techniques
- Image Processing Techniques and Applications
- Advanced Vision and Imaging
- Image and Signal Denoising Methods
- Speech and Audio Processing
- Music and Audio Processing
- Indoor and Outdoor Localization Technologies
- Advanced Image and Video Retrieval Techniques
- Image Retrieval and Classification Techniques
- Robotics and Sensor-Based Localization
- Video Analysis and Summarization
- Video Surveillance and Tracking Methods
- Digital Media Forensic Detection
- Medical Imaging Techniques and Applications
- Medical Image Segmentation Techniques
- Photoacoustic and Ultrasonic Imaging
École Normale Supérieure - PSL
2010-2014
Microsoft (United States)
2014
Institut national de recherche en informatique et en automatique
2009-2011
Centre National de la Recherche Scientifique
2010-2011
École Normale Supérieure
2011
Blur from camera shake is mostly due to the 3D rotation of camera, resulting in a blur kernel that can be significantly non-uniform across image. However, most current deblurring methods model observed image as convolution sharp with uniform kernel. We propose new parametrized geometric blurring process terms rotational velocity during exposure. apply this two different algorithms for removal: first one uses single blurry (blind deblurring), while second both and but noisy same scene. show...
We present a method to replace user specified target region of photograph by using other photographs the same scene downloaded from Internet via viewpoint invariant image search.Each retrieved images is first geometrically then photometrically registered with query photograph.Geometric registration achieved multiple homographies and photometric performed global affine transformation on intensities.Each proposes possible solution for region.In final step we combine these proposals into single...
We address the problem of deblurring images degraded by camera shake blur and saturated or over-exposed pixels. Saturated pixels are a for existing non-blind algorithms because they violate assumption that Image formation process is linear, often cause significant artifacts in deblurred outputs. propose for- ward model includes sensor saturation, use it to derive algorithm properly treating By using this forward reasoning about causes results, we obtain significantly better results than...
Active speaker detection (ASD) and virtual cinematography (VC) can significantly improve the experience of a video conference by automatically panning, tilting zooming camera: subjectively users rate an expert cinematographer higher than unedited video. We describe new automated ASD VC that performs within 0.3 MOS based on subjective ratings with 1-5 scale. This system uses 4K wide-FOV camera, depth microphone array, extracts features from each modality trains using AdaBoost machine learning...
Active speaker detection (ASD) and virtual cinematography (VC) can significantly improve the remote user experience of a video conference by automatically panning, tilting zooming conferencing camera: users subjectively rate an expert cinematographer's higher than unedited video. We describe new automated ASD VC that performs within 0.3 MOS cinematographer based on subjective ratings with 1-5 scale. This system uses 4K wide-FOV camera, depth microphone array; it extracts features from each...