A collection of datasets I've contributed to or collected through research collaborations. These datasets span computer vision, robotics, multimedia, and audio-visual processing. Each dataset includes access links, detailed documentation, and citation information for researchers and practitioners.
CORSMAL Containers Manipulation
A dataset with 1,140 audio-visual-inertial recordings of people interacting with 15 containers, using 4 cameras (RGB, depth, and infrared) and an 8-element circular microphone array. Containers are either empty or filled at 2 different levels (50%, 90%) with 3 different types of content (water, pasta, rice). Includes activities like pouring liquids in glasses/cups and shaking food boxes.
📊 Access Dataset
📋 BibTeX Citation
@misc{Xompero2020CCM,
title = {CORSMAL Containers Manipulation},
author = {Xompero, Alessio and Sanchez-Matilla, Ricardo and Mazzon, Riccardo and Cavallaro, Andrea},
organization={Queen Mary University of London},
doi={https://doi.org/10.17636/101CORSMAL1},
year = {2020}
}
CORSMAL Containers
A vision dataset of 23 transparent and non-transparent containers, such as drinking cups, drinking glasses, and bottles. Each container is placed on a table in two different setups with varying lighting and background conditions. Observed by two cameras with wide-baseline configuration, acquiring RGB, depth, and narrow-baseline stereo infrared images.
📊 Access Dataset
📋 BibTeX Citation
@misc{Xompero2020Containers,
title = {CORSMAL Containers},
author = {Xompero, Alessio and Sanchez-Matilla, Ricardo and Mazzon, Riccardo and Cavallaro, Andrea},
organization={Queen Mary University of London},
doi={https://doi.org/10.17636/corsmal2},
year = {2020}
}
Crop - CORSMAL Containers Manipulation
10,216 RGB images automatically sampled from the three fixed views of the public videos recordings of the CORSMAL Container Manipulation dataset, and capturing cups (4) and drinking glasses (4) as containers under different lighting and background conditions. Containers are completely visible or occluded by the person's hand.
📊 Access Dataset (Zenodo)
📋 BibTeX Citation
@misc{Sanchez-Matilla2021C-CCM,
title = {Crop - CORSMAL Containers Manipulation (C-CCM) (Version 1.0)},
author = {Sanchez-Matilla, Ricardo and Xompero, Alessio and Modas, Apostolos and Frossard, Pascal and Cavallaro, Andrea},
doi={ https://doi.org/10.5281/zenodo.4642577},
year = {2021}
}
CORSMAL Hand-Occluded Containers (CHOC)
An image-based dataset for category-level 6D object pose and size estimation with 138,240 pseudo-realistic composite RGB-D images of hand-held containers on top of 30 real backgrounds (mixed-reality set) and 3,951 RGB-D images selected from the CORSMAL Container Manipulation dataset.
📊 Access Dataset (Zenodo)
📋 BibTeX Citation
@misc{Weber2023CHOC,
title = {CORSMAL Hand-Occluded Containers (CHOC) (Version 1.4)},
author = {Weber, Xavier and Apicella, Tommaso and Xompero, Alessio and Cavallaro, Andrea},
doi={ https://doi.org/10.5281/zenodo.5085800},
year = {2023}
}
X-View: Cross-Camera View Overlap Recognition
A dataset for cross-camera view overlap recognition and spatial alignment. Contains multi-view camera recordings from multiple cameras placed at different positions and orientations. Useful for applications in distributed camera networks, surveillance systems, and multi-camera tracking.
CAV3D: Audio-Visual 3D Speaker Tracking
A dataset for 3D speaker tracking using co-located audio-visual sensing. Collected with a monocular color camera paired with an 8-element circular microphone array (96 kHz sampling). Video at 15 FPS. Includes synchronized data from 4 hardware-triggered CCD cameras at room corners. Supports up to 3 simultaneous speakers in a calibrated, annotated environment.
📊 Access Dataset
📋 BibTeX Citation
@article{Qian2019TMM,
title = {Multi-speaker tracking from an audio-visual sensing device},
author = {Qian, X. and Brutti, A. and Omologo, L. and Cavallaro, Andrea},
journal={IEEE Transactions on Multimedia},
volume={21},
number={10},
month=mar,
year = {2019}
}
MediaEval SEM 2015: Multi-User Event Media
Three event datasets collected for the Synchronization of Multi-User Event Media task at MediaEval 2015. Tour de France 2014: 33 galleries of images from Flickr covering the entire competition. NAMM 2015: 420 Flickr images and 32 YouTube videos in 19 galleries from the Los Angeles event. Spring Party Salesiani 2015: Videos and pictures captured by attendees during a students' party in Trento, Italy.
MediaEval SEM 2014: Olympic Games Datasets
Two Olympic Games image datasets from the Synchronization of Multi-User Event Media task at MediaEval 2014. London Olympics 2012: 2,124 images in 37 galleries capturing the Olympic Games. Vancouver Winter Olympics 2010: 1,351 images in 35 galleries representing various competitions. All images gathered from Flickr and made available under Creative Commons license for research purposes.