Cookies Policy
The website need some cookies and similar means to function. If you permit us, we will use those means to collect data on your visits for aggregated statistics to improve our service. Find out More
Accept Reject
  • Menu
About
Download Photo HD

About

Isabel Rio-Torto received the master's degree in Electrical and Computers Engineering in 2019 from the Faculty of Engineering of the University of Porto (FEUP). Isabel is currently a research assistant at INESC TEC, associated with the Visual Computing and Machine Intelligence Group (VCMI), and a Ph.D. student in Computer Science from the Faculty of Sciences of the University of Porto (FCUP). Isabel is also an Invited Teaching Assistant at FEUP, teaching programming courses. Her work is currently focused on "Self-explanatory computer-aided diagnosis with limited supervision".

Interest
Topics
Details

Details

Publications

2022

From Captions to Explanations: A Multimodal Transformer-based Architecture for Natural Language Explanation Generation

Authors
Torto, IR; Cardoso, JS; Teixeira, LF;

Publication
Pattern Recognition and Image Analysis - 10th Iberian Conference, IbPRIA 2022, Aveiro, Portugal, May 4-6, 2022, Proceedings

Abstract

2022

Hybrid Quality Inspection for the Automotive Industry: Replacing the Paper-Based Conformity List through Semi-Supervised Object Detection and Simulated Data

Authors
Rio-Torto, I; Campanico, AT; Pinho, P; Filipe, V; Teixeira, LF;

Publication
APPLIED SCIENCES-BASEL

Abstract
The still prevalent use of paper conformity lists in the automotive industry has a serious negative impact on the performance of quality control inspectors. We propose instead a hybrid quality inspection system, where we combine automated detection with human feedback, to increase worker performance by reducing mental and physical fatigue, and the adaptability and responsiveness of the assembly line to change. The system integrates the hierarchical automatic detection of the non-conforming vehicle parts and information visualization on a wearable device to present the results to the factory worker and obtain human confirmation. Besides designing a novel 3D vehicle generator to create a digital representation of the non conformity list and to collect automatically annotated training data, we apply and aggregate in a novel way state-of-the-art domain adaptation and pseudo labeling methods to our real application scenario, in order to bridge the gap between the labeled data generated by the vehicle generator and the real unlabeled data collected on the factory floor. This methodology allows us to obtain, without any manual annotation of the real dataset, an example-based F1 score of 0.565 in an unconstrained scenario and 0.601 in a fixed camera setup (improvements of 11 and 14.6 percentage points, respectively, over a baseline trained with purely simulated data). Feedback obtained from factory workers highlighted the usefulness of the proposed solution, and showed that a truly hybrid assembly line, where machine and human work in symbiosis, increases both efficiency and accuracy in automotive quality control.

2021

Automatic quality inspection in the automotive industry: a hierarchical approach using simulated data

Authors
Rio-Torto, I; Campanico, AT; Pereira, A; Teixeira, LF; Filipe, V;

Publication
2021 IEEE 8th International Conference on Industrial Engineering and Applications (ICIEA)

Abstract

2021

Improving Automatic Quality Inspection in the Automotive Industry by Combining Simulated and Real Data

Authors
Pinho, P; Rio Torto, I; Teixeira, LF;

Publication
ADVANCES IN VISUAL COMPUTING (ISVC 2021), PT I

Abstract
Considerable amounts of data are required for a deep learning model to generalize to unseen cases successfully. Furthermore, such data is often manually labeled, making its annotation process costly and time-consuming. We propose using unlabeled real-world data in conjunction with automatically labeled synthetic data, obtained from simulators, to surpass the increasing need for annotated data. By obtaining real counterparts of simulated samples using CycleGAN and subsequently performing fine-tuning with such samples, we manage to improve a vehicle part’s detection system performance by 2.5%, compared to the baseline exclusively trained on simulated images. We explore adding a semantic consistency loss to CycleGAN by re-utilizing previous work’s trained networks to regularize the conversion process. Moreover, the addition of a post-processing step, which we denominate global NMS, highlights our approach’s effectiveness by better utilizing our detection model’s predictions and ultimately improving the system’s performance by 14.7%. © 2021, Springer Nature Switzerland AG.

2020

Understanding the decisions of CNNs: an in-model approach

Authors
Rio Torto, I; Fernandes, K; Teixeira, LF;

Publication
PATTERN RECOGNITION LETTERS

Abstract

Supervised
thesis

2021

Combining simulated and real images in deep learning

Author
Pedro Xavier Tavares Monteiro Correia de Pinho

Institution
UP-FEUP