- Advanced Image and Video Retrieval Techniques
- Image Retrieval and Classification Techniques
- Video Surveillance and Tracking Methods
- Advanced Vision and Imaging
- Video Analysis and Summarization
- Face and Expression Recognition
- Face recognition and analysis
- Human Pose and Action Recognition
- Visual Attention and Saliency Detection
- Robotics and Sensor-Based Localization
- Anomaly Detection Techniques and Applications
- Topic Modeling
- Emotion and Mood Recognition
- Image Processing Techniques and Applications
- Supply Chain and Inventory Management
- Medical Image Segmentation Techniques
- Data Management and Algorithms
- Natural Language Processing Techniques
- Image Enhancement Techniques
- Advanced Image Processing Techniques
- Image and Signal Denoising Methods
- Hand Gesture Recognition Systems
- Context-Aware Activity Recognition Systems
- Text and Document Classification Technologies
- Music and Audio Processing
University of Palermo
2016-2025
Boston University
1997-2003
TSI (United States)
1986
A technique for 3D head tracking under varying illumination is proposed. The modeled as a texture mapped cylinder. Tracking formulated an image registration problem in the cylinder's map image. resulting dynamic provides stabilized view of face that can be used input to many existing 2D techniques recognition, facial expressions analysis, lip reading, and eye tracking. To solve with lighting variation motion, residual error linear combination warping templates orthogonal templates. Fast...
ImageRover is a search by image content navigation tool for the world wide web. To gather images expediently, collection subsystem utilizes distributed fleet of WWW robots running on different computers. The information about they find, computing appropriate decompositions and indices, store this extracted in vector form searches based content. At time, users can iteratively guide through selection relevant examples. Search performance made efficient use an approximate, optimized k-d tree...
A system is proposed that combines textual and visual statistics in a single index vector for content-based search of WWW image database. Textual are captured form using latent semantic indexing (LSI) based on text the containing HTML document. Visual color orientation histograms. By an integrated approach, it becomes possible to take advantage statistical couplings between content document (latent content) contents images (visual statistics). The combined approach allows improved...
This paper presents a novel and fully automatic technique to estimate depth information from single input image. The proposed method is based on new image classification able classify digital images (also in Bayer pattern format) as indoor, outdoor with geometric elements or without elements. Using the collected step suitable map estimated. unsupervised generate view of scene, requiring low computational resources.
With the advent of modern pre-trained Transformers, text preprocessing has started to be neglected and not specificly addressed in recent NLP literature. However, both from a linguistic computer science point view, we believe that even when using can significantly impact on performance classification model. We want investigate compare, through this study, how impacts Text Classification (TC) traditional models. report discuss techniques found literature their most variants or applications...
A system is proposed that combines textual and visual statistics in a single index vector for content-based search of WWW image database. Textual are captured form using latent semantic indexing based on text the containing HTML document. Visual color orientation histograms. By an integrated approach, it becomes possible to take advantage statistical couplings between content document (latent content) contents images (visual statistics). The combined approach allows improved performance...
Guided by a corpus linguistics approach, in this article we present comparative evaluation of State-of-the-Art (SotA) models, with special focus on Transformers, to address the task Fake News Spreaders (i.e., users that share News) detection. First, explore reference multilingual dataset for considered task, exploiting techniques, such as chi-square test, keywords and Word Sketch. Second, perform experiments several models Natural Language Processing. Third, using most recent...
The increasing development of advanced multimedia applications requires new technologies for organizing and retrieving by content databases still digital images or video sequences. authors describe JACOB, a prototypal system allowing content-based browsing querying in databases. JACOB automatically splits into sequence shots, extracts few representative frames (said r-frames) from each shot computes r-frame descriptors based on features like color texture. No user action is required during...
This work presents a new unsupervised technique aimed to generate stereoscopic views estimating depth information from single input image. Using image, vanishing lines/points are extracted using few heuristics an approximated map. The map is then used stereo pairs. overall method well suited for real time application and works also on CFA (colour filtering array) data acquired by consumer imaging devices. Experimental results large dataset reported.
The advent of deepfake technology has raised significant concerns regarding its impact on individuals' cognitive processes and beliefs, considering the pervasive relationships between human cognition. This study delves into psychological literature surrounding deepfakes, focusing people's public representation this emerging highlighting prevailing themes, opinions, emotions. Under media framing, theoretical framework is crucial in shaping schemas technology. A qualitative method been applied...
An improved technique for 3D head tracking under varying illumination conditions is proposed. The modeled as a texture mapped cylinder. Tracking formulated an image registration problem in the cylinder's map image. To solve presence of lighting variation and motion, residual error linear combination warping templates orthogonal templates. Fast stable on-line then achieved via regularized weighted least squares minimization error. regularization term tends to limit potential ambiguities that...
This work presents a new unsupervised technique aimed to generate stereoscopic views estimating depth information from single input image. Using image, vanishing lines/points are extracted using few heuristics an approximated map. The map is then used stereo pairs. overall method well suited for real time application and works also on CFA (colour filtering array) data acquired by consumer imaging devices. Experimental results large dataset reported.
In the last years a lot of work has been done on color, textural, structural and semantic indexing "content-based" video databases. Motion-based less explored, with approaches generally based analysis optical flows. Compressed videos require decompression sequences computation flows, two steps computationally heavy. this paper we propose some methods to index by motion features (mainly related camera motion) motion-based spatial segmentation frames, in fully automatic way. Our idea is use...
The increasing development of advanced multimedia applications requires new technologies for organizing and retrieving by content databases digital video. Several based features (color, texture, motion, etc.) are needed to perform a reliable retrieval. We present method automatic motion video indexing A prototypal system has been developed prove the validity our approach. Our automatically splits into sequence shots, extracts few representative frames (said r-frames) from each shot computes...
A novel method for 3D head tracking in the presence of large rotations and facial expression changes is described. Tracking formulated terms color image registration texture map a surface model. Model appearance recursively updated via mosaicking as orientation varies. The resulting dynamic provides stabilized view face that can be used input to many existing 2D techniques recognition, expressions analysis, lip reading, eye tracking. Parameters are estimated robust minimization procedure;...