1 code implementation • NeurIPS 2023 • Thalles Silva, Adín Ramírez Rivera
We extensively ablate our method and demonstrate that our proposed random partition pretext task improves the quality of the learned representations by devising multiple random classification tasks.
no code implementations • 3 Oct 2023 • Bruno Souza, Marius Aasan, Helio Pedrini, Adín Ramírez Rivera
Scene graphs (SGs) have emerged as a useful tool for multimodal image analysis, showing impressive performance in tasks such as Visual Question Answering (VQA).
1 code implementation • 1 Oct 2023 • Thalles Santos Silva, Helio Pedrini, Adín Ramírez Rivera
We present Contextualized Local Visual Embeddings (CLoVE), a self-supervised convolutional-based method that learns representations suited for dense prediction tasks.
1 code implementation • 19 Jul 2022 • Darwin Saire, Adín Ramírez Rivera
Our proposed model addresses this problem by providing an internal structure for the feature representations while extracting a global representation that supports the former.
no code implementations • 13 Jul 2022 • Patrik Joslin Kenfack, Kamil Sabbagh, Adín Ramírez Rivera, Adil Khan
Fairness has become an essential problem in many domains of Machine Learning (ML), such as classification, natural language processing, and Generative Adversarial Networks (GANs).
no code implementations • 31 May 2022 • Sandra Robles, Jonathan S. Gómez, Adín Ramírez Rivera, Nelson D. Padilla, Diego Dujovne
A key ingredient for semi-analytic models (SAMs) of galaxy formation is the mass assembly history of haloes, encoded in a tree structure.
1 code implementation • 31 Dec 2021 • Thalles Silva, Adín Ramírez Rivera
We introduce Consistent Assignment for Representation Learning (CARL), an unsupervised learning method to learn visual representations by combining ideas from self-supervised contrastive learning and deep clustering.
1 code implementation • 28 May 2021 • Darwin Saire, Adín Ramírez Rivera
The semantic segmentation (SS) task aims to create a dense classification by labeling at the pixel level each object present on images.
no code implementations • 2 Apr 2021 • Miguel Rodríguez Santander, Juan Hernández Albarracín, Adín Ramírez Rivera
In this paper, we explore the effects of stacking databases, model initialization, and data amplification techniques when training with limited data on deep learning models' performance.
no code implementations • EACL (VarDial) 2021 • Albina Khusainova, Adil Khan, Adín Ramírez Rivera, Vitaly Romanov
The choice of parameter sharing strategy in multilingual machine translation models determines how optimally parameter space is used and hence, directly influences ultimate translation quality.
no code implementations • ICLR Workshop EBM 2021 • Thalles Santos Silva, Adín Ramírez Rivera
An unsupervised learning method to learn visual representations by combining contrastive learning with deep clustering.
1 code implementation • 30 Jan 2021 • Juan F. Hernández Albarracín, Adín Ramírez Rivera
Experiments on video reenactment show the effectiveness of our disentanglement in the input space where our model outperforms the baselines in reconstruction quality and motion alignment.
no code implementations • 10 Oct 2020 • Adín Ramírez Rivera, Adil Khan, Imad E. I. Bekkouch, Taimoor S. Sheikh
Anomaly detection suffers from unbalanced data since anomalies are quite rare.
1 code implementation • 23 Feb 2020 • Rodolfo Quispe, Darwin Ttito, Adín Ramírez Rivera, Helio Pedrini
Crowd scene analysis has received a lot of attention recently due to the wide variety of applications, for instance, forensic science, urban planning, surveillance and security.
no code implementations • 22 Jun 2019 • Sandra Robles, Jonathan S. Gómez, Adín Ramírez Rivera, Jenny A. González, Nelson D. Padilla, Diego Dujovne
Our aim is to provide a new framework for halo merger tree generation that takes advantage of the results of large volume simulations, with a modest computational cost.
1 code implementation • 29 May 2019 • Darwin Saire Pilco, Adín Ramírez Rivera
We propose the Graph Learning Network (GLN), a simple yet effective process to learn node embeddings and structure prediction functions.
1 code implementation • 31 Mar 2019 • Albina Khusainova, Adil Khan, Adín Ramírez Rivera
We evaluate state-of-the-art word embedding models for two languages using our proposed datasets for Tatar and the original datasets for English and report our findings on performance comparison.