Alessio Tonioni

Alessio Tonioni

Research Scientist Manager @ Google.

Posts by Collection

portfolio

publications

Learning confidence measures in the wild

Authors: Fabio Tosi, Matteo Poggi, Alessio Tonioni, Luigi Di Stefano, & Stefano Mattoccia
Published in 28th British Machine Vision Conference (BMVC 2017), 2017

In this paper we propose a methodology suited for training a confidence measure for stereo in a self-supervised manner.

Download here

Unsupervised Adaptation For Deep Stereo

Authors: Alessio Tonioni, Matteo Poggi, stefano Mattoccia and Luigi Di Stefano
Published in Proceedings of the IEEE International Conference on Computer Vision, 2017

In this paper we propose a novel unsupervised adaptation approach that enables to fine-tune a deep learning stereo model without any ground-truth information.

Download here

Learning to detect good 3D keypoints

Authors: Alessio Tonioni, Samuele Salti, Federico Tombari, Riccardo Spezialetti and Luigi Di Stefano
Published in International Journal of Computer Vision, 2018

In this paper we learn a descriptor-specific 3D keypoint detector so as to optimize the end-to-end performance of a feature matching pipeline

Download here

A deep learning pipeline for product recognition in store shelves

Authors: Alessio Tonioni, Eugenio Serra and Luigi Di Stefano.
Published in International Conference on Image Processing, Applications and Systems, 2018

In this paper, we propose a deep learning pipeline to recognize products on grocery shelves that can scale effortlessly to thousand of different products to recognize.

Download here

Exploiting Semantics in Adversarial Training for Image-Level Domain Adaptation

Authors: Pierluigi Zama Ramirez, Alessio Tonioni and Luigi Di Stefano
Published in International Conference on Image Processing, Applications and Systems, 2018

In this paper, we address the problem of domain adaptation for computer vision by learning a domain-to-domain image translation GAN. Peculiarly to our method, we introduce semantic constraints into the generation process to both avoid artifacts and guide the synthesis

Download here

Real-Time Highly Accurate Dense Depth on a Power Budget using an FPGA-CPU Hybrid SoC

Authors: Oscar Rahnama, Tommaso Cavallari, Stuart Golodetz, Alessio Tonioni, Thomas Joy, Luigi Di Stefano, Simon Walker, and Philip HS Torr
Published in IEEE Transactions on Circuits and Systems II: Express Briefs , 2019

In this paper, we leverage a FPGA-CPU chip to propose a novel, sophisticated, stereo approach that combines the best features of SGM and ELAS-based methods to compute highly accurate dense depth in real time.

Download here

Learning to Adapt for Stereo

Authors: Alessio Tonioni, Oscar Rahnama, Thomas Joy, Luigi Di Stefano, Thalaiyasingam Ajanthan, and Philip HS Torr.
Published in Conference on Computer Vision and Pattern Recognition, 2019

In this paper we introduce a” learning-to-adapt” framework that enables deep stereo methods to continuously adapt to new target domains in an unsupervised manner.

Download here

Real-time self-adaptive deep stereo

Authors: Alessio Tonioni, Fabio Tosi, Matteo Poggi, Stefano Mattoccia and Luigi Di Stefano
Published in Conference on Computer Vision and Pattern Recognition, 2019

In this paper we propose a real-time self adaptive deep stereo system.

Download here

Semi-Automatic Labeling for Deep Learning in Robotics

Authors: Daniele De Gregorio, Alessio Tonioni, Gianluca Palli and Luigi Di Stefano
Published in IEEE Transactions on Automation Science and Engineering , 2019

In this paper, we propose Augmented Reality Semi-automatic labeling (ARS), a semi-automatic method to create large labeled datasets with minimal human intervention

Download here

Learning Across Tasks and Domains

Authors: Pierluigi Zama Ramirez, Alessio Tonioni, Samuele Salti and Luigi Di Stefano
Published in International Conference on Computer Vision, 2019

In this work, we introduce a novel adaptation framework that can operate across both task and domains.

Download here

Unsupervised Domain Adaptation for Depth Prediction from Images

Authors: Alessio Tonioni, Matteo Poggi, Stefano Mattoccia and Luigi Di Stefano
Published in IEEE Transactions on Pattern Analysis and Machine Intelligence, 2019

In this paper we extend our previous unsupervised adaptation approach to fine-tune a deep learning stereo or mono model without any ground-truth information.

Download here

Continual Adaptation for Deep Stereo

Authors: Matteo Poggi, Alessio Tonioni, Fabio Tosi, Stefano Mattoccia and Luigi Di Stefano
Published in IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020

In this paper we propose an etension of our real-time self adaptive deep stereo system.

Download here

talks

teaching

Academic Tutor

Undergraduate course, University of Bologna, computer engineering course, 2019

I have been academic tutor for the courses:

  • Computer Vision and Image Processing M - 2016/2017
  • Computer Vision and Image Processing M - 2017/2018
  • Reti Logiche T - 2019/2020

Corporate Trainer

Corporate Training Course, Fondazione Aldini Valeriani, 2019

Corporate Trainer for Fondazione Aldini Valeriani on Machine-Deep Learning.