A5033711815- Advanced Image and Video Retrieval Techniques
- Video Analysis and Summarization
- Image Retrieval and Classification Techniques
- Multimodal Machine Learning Applications
- Image Processing and 3D Reconstruction
- Data Management and Algorithms
- Robotics and Sensor-Based Localization
- Digital Rights Management and Security
- Music and Audio Processing
- Remote-Sensing Image Classification
- Satellite Image Processing and Photogrammetry
- Open Education and E-Learning
- Geochemistry and Geologic Mapping
- Hydrocarbon exploration and reservoir analysis
- Cultural Heritage Management and Preservation
- Multimedia Communication and Technology
- Advanced Malware Detection Techniques
- COVID-19 diagnosis using AI
- Chaos-based Image/Signal Encryption
- Web Data Mining and Analysis
- Service and Product Innovation
- Big Data and Business Intelligence
- Music Technology and Sound Studies
- Smart Parking Systems Research
- Optical Wireless Communication Technologies
Istituto di Scienza e Tecnologie dell'Informazione "Alessandro Faedo"
2015-2024
National Research Council
2018-2024
Institute of Scientific and Technical Information of China
2013-2022
Italian Resuscitation Council
2020
Confederazione Nazionale dell'Artigianato e Della Piccola e Media Impresa
2016
National Academies of Sciences, Engineering, and Medicine
2011
The scalability, as well the effectiveness, of different Content-based Image Retrieval (CBIR) approaches proposed in literature, is today an important research issue. Given wealth images on Web, CBIR systems must fact leap towards Web-scale datasets. In this paper, we report our experience building a test collection 100 million images, with corresponding descriptive features, to be used experimenting new scalable techniques for similarity searching, and comparing their results. context SAPIR...
Despite the fact that automatic content analysis has made remarkable progress over last decade - mainly due to significant advances in machine learning interactive video retrieval is still a very challenging problem, with an increasing relevance practical applications. The Video Browser Showdown (VBS) annual evaluation competition pushes limits of state-of-the-art tools, tasks, data, and metrics. In this paper, we analyse results outcome 8th iteration VBS detail. We first give overview novel...
Content-Based Image Retrieval in large archives through the use of visual features has become a very attractive research topic recent years. The cause this strong impulse area is certainly to be attributed Convolutional Neural Network (CNN) activations as and their outstanding performance. However, practically all available image retrieval systems are implemented main memory, limiting applicability preventing usage big-data applications. In paper, we propose transform CNN into textual...
This paper describes in detail VISIONE, a video search system that allows users to for videos using textual keywords, the occurrence of objects and their spatial relationships, colors image similarity. These modalities can be combined together express complex queries meet users’ needs. The peculiarity our approach is we encode all information extracted from keyframes, such as visual deep features, tags, color object locations, convenient encoding indexed single text retrieval engine. offers...
An increasing number of people share their thoughts and the images lives on social media platforms. People are exposed to food in everyday on-line what they eating by means photos taken dishes. The hashtag #foodporn is constantly among popular hashtags Twitter second most subject Instagram after selfies. system that we propose, WorldFoodMap, captures stream from and, thanks a CNN image classifier, identifies categories sharing. By collecting associating category location them, WorldFoodMap...
In this paper, the performance of several visual features is evaluated in automatically recognizing landmarks (monuments, statues, buildings, etc.) pictures. A number were selected for test. Pictures taken from a test set classified trying to guess which landmark they contained. We both global and local features. As expected, performed better given their capability being less affected variations that are mainly static objects generally also maintain Between features, SIFT outperformed SURF ColorSIFT.
In this paper we present a Wireless Sensor Network (WSN), which is intended to provide scalable solution for active cooperative monitoring of wide geographical areas. The system designed use different smart-camera prototypes: where the connection power grid available powerful embedded hardware implements Deep Neural Network, otherwise fully autonomous energy-harvesting node based on low-energy custom board employs lightweight image analysis algorithms. Parking lots occupancy in historical...
In this paper we propose a novel approach that allows processing image content based queries expressed as arbitrary combinations of local and global visual features, by using single index realized an inverted file. The was implemented on top the Lucene retrieval engine. This is particularly useful to allow people efficiently interactively check quality result exploiting various features when usingvarious systems.
In this paper we present the architecture of a Digital Library for enabling reusing audiovisual documents in an e-Learning context. The reuse Learning Objects is based on automatically extracted descriptors carrying semantic meaning professional that uses these to prepare new interactive multimedia lectures. presented system MILOS, general purpose Multimedia Content Management System created support design and effective implementation digital library applications. MILOS supports storage...
VISIONE is a versatile video retrieval system supporting diverse search functionalities, including free-text, similarity, and temporal searches. Its recent success in securing first place the 2024 Video Browser Showdown (VBS) highlights its effectiveness. Originally designed for analyzing, indexing, searching content, can also be adapted to images from lifelog cameras thanks reliance on frame-based representations mechanisms.
VISIONE is a large-scale video retrieval system that integrates multiple search functionalities, including free text search, spatial color and object visual semantic similarity temporal search. The leverages cutting-edge AI technology for analysis advanced indexing techniques to ensure scalability. As demonstrated by its runner-up position in the 2023 Video Browser Showdown competition, effectively these capabilities provide comprehensive solution. A demo available online, showcasing on over...
In this paper we present a prototype for parental control that detects images with adult content received on mobile device. More specifically, the application developed is able to intercept through various communication channels (bluetooth, MMS) devices based Symbian™ operating systems. Once intercepted, are analysed by component of system automatically classify explicit sexual content. At current stage runs device, classifier remote server.