Creating and sharing knowledge for telecommunications

Project: Compression of Multimodal Biomedical Images using Neural Networks

Acronym: CoMBINNe
Main Objective:
The main goals of this research plan are: (i) to investigate and propose new prediction modes to increase the multimodal compression efficiency of standard
encoders, based on image to image (I2I) translation-based methods and cross-modality prediction; (ii) to investigate the coding efficiency of end-to-end learningbased
multi-modal image codecs using state-of-the-art learning-based encoders to devise novel techniques capable of surpass standard encoders. Overall, the
project seeks to explore multi-dimensional and multimodal characteristics to achieve better compression rates than current standards.
Reference: 2022.09914.PTDC
Funding: FCT
Approval Date: 09-12-2022
Start Date: 12-03-2023
End Date: 11-03-2025
Team: Lucas Arrabal Thomaz, Pedro Antonio Amado Assunção, Luís Miguel de Oliveira Pegado de Noronha e Távora, Sérgio Manuel Maciel de Faria, Daniel Filipe da Silva Nicolau, Nicolas David Freire Vasconcellos, Eduardo António Barros da Silva, João Oliveira Parracho
Groups: Multimedia Signal Processing – Lr
Partners: Universidade Federal do Rio de Janeiro
Local Coordinator: Lucas Arrabal Thomaz
...

Associated Publications
  • 2Papers in Conferences
  • O. Parracho, E. Silva, L. A. Thomaz, L.M. Távora, S.M.M. Faria, Non-Separablewavelet Transform Using Learnable Convolutional Lifting Steps, IEEE International Conference on Image Processing ICIP, Abu Dhabi, United Arab Emirates, Vol., pp. -, October, 2024 | BibTex
  • D. Nicolau, O. Parracho, L. A. Thomaz, L.M. Távora, S.M.M. Faria, Enhanced residue prediction for Lossless coding of multi- modal image pairs based on image to image translation, IEEE European Workshop on Visual Information Processing - EUVIP, Gjøvik, Norway, Vol., pp. -, September, 2023,
    | Abstract
    | BibTex