• logo_github2
  • logo_gmail3

©2021 by Gwena Cunha

ABOUT

I'm a Brazilian computer engineer who recently obtained her Ph.D. in Electronics Engineering (focus on Machine Learning) in South Korea. During my graduate studies, I have gained expertise in the areas of Robust AI, Natural Language Processing / Understanding / Generation with noisy data, and Multimodal data. I have enjoyed presenting my work at various conferences such as COLING and ICML and getting valuable feedback through discussions with other presenters. I'm passionate about democratizing AI by helping the community in open-source projects (TLM, AI for COVID-19) and by sharing my personal codes.

 

CODING

NATURAL LANGUAGE PROCESSING

EA & SOLAR CELLS

image_solar_cell_code_edited.png

MULTIMODAL

coding_multimodal_2_resized.png

MACHINE LEARNING NOTEBOOKS

DATASETS

 

PROJECTS

Workspace

PAPER NOTES

logo_tlm.png

THE LEARNING MACHINE

Open-Source Project: NLP in Machine Learning Team Leader, Programmer and Writer

dll_logo.png

DEEP LEARNING LABS

Time series prediction for COVID-19

NASA SPACE APPS CHALLENGE

"Mapping the environmental benefits of COVID-19 lockdowns: A lesson for combating air pollution"

 

PUBLICATIONS

2020

Cunha Sergio, G., Moirangthem, D. S., and Lee, M. (2020). “Attentively Embracing Noise for Robust Latent Representation in BERT”. In: Proceedings of the 38th International Conference on Computational Linguistics (COLING 2020), Barcelona, Spain (Online), 8-13 December 2020. ACL, pp. 3479-3491. [paper] [code]

Sergio, G. C. and Lee, M. (2020). "Scene2Wav: a Deep Convolutional Sequence-to-Conditional SampleRNN for Emotional Scene Musicalization". In: Multimedia Tools and Applications 2020. Sep 2020. Springer, pp. 1-20. [paper] [code] [KNU-ICT Research Award]

Sergio, G. C. and Lee, M. (2020). “Robust Sentiment Classification in Noisy User-generated Text with BERT-based model”. In: The 37th International Conference on Machine Learning (ICML 2020). The 1st Women in Machine Learning un-Workshop. Virtual, 13-18 July 2020. [slideslive] [code]

 

Sergio, G. C. and Vincent, P. et al. (2020). “Application of Genetic Algorithm for More Efficient Multi-layer Thickness Optimization in Solar Cells”. In: Energies 13 (7). April 2020. MDPI, pp1-14. [paper] [code]

Sergio, G. C. and Lee, M. (2020). “Emotional Video to Audio Transformation Using Deep Recurrent Neural Networks and a Neuro-Fuzzy System”. In: Mathematical Problems in Engineering 2020. Feb 2020. Hindawi, pp. 1-15. [paper] [code]

2019

Sergio, G. C. and Vincent, P. et al. (2019). “Alternative Approach to Optimizing Optical Spacer Layer Thickness in Solar Cell Using Evolutionary Algorithm”. In: Proceedings of the 19th International Conference on Numerical Simulation of Optoelectronic Devices (NUSOD 2019), Ottawa, Canada, 8-12 July 2019. IEEE, pp. 89-90. [paper] [code]

2018

Sergio, G. C. and Lee, M. (2018). "Audio Generation with Deep Recurrent Neural Networks". In: 18th China-Japan-Korea Joint Workshop on Neurobiology and Neuroinformatics (NBNI 2018). Jeju-do, South Korea, Oct 2018.

Sergio, G. C. and Moirangthem, D. S. and Lee, M. (2018). “Temporal Hierarchies in Sequence to Sequence for Sentence Correction”. In: 2018 International Joint Conference on Neural Networks (IJCNN 2018), Rio de Janeiro, Brazil, 08-13 July 2018. IEEE, pp. 1–7. [ppt] [paper] [code]

Sergio, G. C. and Lee, M. (2018). "Deep Recurrent Neural Network applied to Incomplete Data". In: Brain and Artificial Intelligence Symposium (BAIS 2018). Gangwon-do, South Korea, Jan 2018. Brain Engineering Society of Korea (BESK). [Best Poster Award]

2016

Sergio, G. C. and Lee, M. (2016). “Audio Generation from Scene Considering Its Emotion Aspect”. In: Proceedings of the 23rd International Conference on Neural Information Processing (ICONIP 2016). Kyoto, Japan, Oct 2016: Springer, pp. 74–81. [paper]

 

2015

Sergio, G. C. et al. (2015). “Generating Music from an Image”. In: Proceedings of the 3rd International Conference on Human-Agent Interaction (HAI 2015). Daegu, South Korea, Oct 2015: ACM, pp. 213–216. [paper]

2021

Cunha Sergio, G. and Lee, M. (2021). “Stacked DeBERT: All Attention in Incomplete Data for Text Classification”. In: Neural Networks 136, pp. 87-96. ISSN:0893-6080. DOI:https://doi.org/10.1016/j.neunet.2020.12.018. [paper] [arxiv] [code]

 

AWARDS

KNU-ICT Research Award

(2019)

For work titled "Scene2Wav: a Deep Convolutional Sequence-to-Conditional SampleRNN for Emotional Scene Musicalization"

Science Without Borders

(2012-2013)

 

Scholarship recipient at Brown University, USA 

BAIS Best Paper Award

(2018)

For work titled "Deep Recurrent Neural Network applied to Incomplete Data"

Valedictorian

(2009-2014)

 

Bachelor of Science: 1st place in university entrance exam and graduating class (Computer Engineering)

KINGS Scholarship

(2014-2020)

 

Recipient of the KNU International Graduate Scholarship

 
 

VOLUNTEERING

ACL'20

On-call volunteer

ICML'20

On-call volunteer,

LXAI Media Liaison

TA'18-'20

Instructor at ABR Lab's Basic Education Program for new students

Peer reviewer

IJCNN'18-'19, ESANN'19, ICANN'19, IJFS'19-'20, NEPL'19

HAI'15

Tour guide

WCCI'18

Session chair: "Neural Models of Perception, Cognition and Action"