Adaptive cross-modal embeddings for image-text alignment

J Wehrmann, C Kolling, RC Barros - … of the AAAI conference on artificial …, 2020‏ - aaai.org
Abstract a using an embedding vector of an instance from modality b. Such an adaptation is
designed to filter and enhance important information across internal features, allowing for …

Language-agnostic visual-semantic embeddings

J Wehrmann, DM Souza, MA Lopes… - Proceedings of the …, 2019‏ - openaccess.thecvf.com
This paper proposes a framework for training language-invariant cross-modal retrieval
models. We also introduce a novel character-based word-embedding approach, allowing …

Bidirectional retrieval made simple

J Wehrmann, RC Barros - Proceedings of the IEEE …, 2018‏ - openaccess.thecvf.com
This paper provides a very simple yet effective character-level architecture for learning
bidirectional retrieval models. Aligning multimodal content is particularly challenging …

Fast query-by-example speech search using attention-based deep binary embeddings

Y Yuan, L **e, CC Leung, H Chen… - IEEE/ACM Transactions …, 2020‏ - ieeexplore.ieee.org
State-of-the-art query-by-example (QbE) speech search approaches usually use recurrent
neural network (RNN) based acoustic word embeddings (AWEs) to represent variable …

Image-text matching using multi-subspace joint representation

H Sun, X Qin, X Liu - Multimedia Systems, 2023‏ - Springer
Joint representation learning has been an attractive way to solve image-text retrieval
problem due to its efficiency on both time and storage. On the one hand, the most classical …

Fast and efficient text classification with class-based embeddings

J Wehrmann, C Kolling… - 2019 International Joint …, 2019‏ - ieeexplore.ieee.org
Current state-of-the-art approaches for Natural Language Processing tasks such as text
classification are either based on Recurrent or Convolutional Neural Networks …

Fully-disentangled text-to-image synthesis

MA Lopes - 2019‏ - tede2.pucrs.br
Algoritmos de generalização sofrem vários problemas. Comumente, os algoritmos de
aprendizado profundo tendem a se ajustar a uma quantidade limitada de dados, gerando …

[PDF][PDF] Style transfer for text-based image manipulation

MD Móre - 2019‏ - repositorio.pucrs.br
Grande parte dos dados que produzimos atualmente estão na forma de fotografias digitais,
o que aumenta a demanda por aplicações de edição de imagens. Contudo, manipulação …