Andrea Cavallaro


Jump to: Recent code from the group | CORSMAL repositories | Web-tools | Datasets

Recent code from the team

Multi-modal emotion recognition: Self-attention and Cross-modal attention models

Training privacy-preserving video analytics pipelines: Suppressing features that reveal private attributes

Conflict-Intensity Estimation: Estimating the level of verbal conflict from speech signals

AV3T: Multi-speaker tracking from a compact audio-visual sensing device

FilterFool: Semantically Adversarial and Learnable Image Processing Filters

ColorFool: Adversarial Image Colorization

Edgefool: An Adversarial Image Enhancement Filter

Robust Private FGSM: Exploiting Vulnerabilities of Deep Neural Networks for Privacy Protection

Scene privacy protection: Adversarial noise that conceals the scene depicted in a picture

PrivEdge: From Local to Distributed Private Training and Prediction

WiSE-MNet: Wireless Simulation Environment for Multimedia Networks

CORSMAL repositories

Benchmark for Human-to-Robot Handovers

The CORSMAL Challenge evaluation toolkit

Towards safe human-to-robot handovers of unknown containers

Audio Classification of the Content of Food Containers and Drinking Glasses

Multi-view shape estimation of transparent containers


PUIQE: a Platform for Underwater Image Quality Evaluation

Social Video: an automated multi-view editing platform

Synch: automatic multi-track synchronisation platform

3Dviz: an interactive trajectory visualisation platform

Recent datasets

CORSMAL Containers Manipulation (audio-visual-inertial data)

CORSMAL Containers (visual data)

CAV3D: Co-located Audio-Visual streams with 3D tracks (audio-visual data)