Remote palpation project:  My responsibilities include defining the project technologies and strategy, leveraging state-of-the-art solutions, providing technical expertise in computer vision techniques and algorithms, leading research and development efforts, managing the project team, ensuring the quality and evaluation of computer vision models, and overseeing project delivery and deployment.   

Elderly surveillance: I investigate the fusion of visual and acoustic data to support the safe life of the elderly living alone. I employ information from visual and acoustic sensors and recognize events, such as domestic activities or abnormal events. In this project, I am developing the following technologies: 1) fast neural networks for extraction of visual characteristics from videos, 2) neural networks for acoustic data analysis and, 3) data fusion for event recognition. This project is on its initial steps, which includes data collection and analysis. Grant website: https://kaken.nii.ac.jp/en/grant/KAKENHI-PROJECT-19K20335/

[3] LS Souza, BB Gatto, JH Xue, K Fukui. “Enhanced Grassmann discriminant analysis with randomized time warping for motion recognition”. PR 2020 (Pattern Recognition, Elsevier)

[4] MAF Mollinetti, MTRS Neto, BB Gatto, T Kuno. “A-DVM: A Self-adaptive Variable Matrix Decision Variable Selection Scheme for Multimodal Problems”. Entropy 2020 (Entropy, MDPI) status: in press.

[8] LS Souza, N Sogi, BB Gatto, K Fukui. “An Interface between Grassmann manifolds and vector spaces”. CVPRW 2020 (International Workshop on Differential Geometry in Computer Vision and Machine Learning)

[9] N Sogi, LS Souza, BB Gatto, K Fukui “Metric Learning with A-based Scalar Product for Image-set Recognition”. CVPRW 2020 (International Workshop on Differential Geometry in Computer Vision and Machine Learning)

Tensor analysis: Here, I investigate and develop a discriminative mechanism to handle multilinear data (e.g., RGB-D, hyperspectral images). By employing the product of spaces, I represent and analyze hidden structures in multilinear data that are not observable through classic methods. The techniques developed in this research are flexible to handle both handcrafted and deep learning representations. The following papers present my findings:

[5] BB Gatto, EM Santos, MAF Molinetti, K Fukui. “Multilinear Clustering via Tensor Fukunaga-Koontz Transform with Fisher Eigenspectrum Regularization”. ASOC 2020 (Applied Soft Computing, Elsevier) status: under review.

[6] BB Gatto, EM Santos, AL Koerich, K Fukui, WSS Júnior. “Tensor Analysis with n-mode Generalized Difference Subspace”. ESWA 2020 (Expert Systems with Applications, Elsevier) status: under review.

[10] BB Gatto, MAF Molinetti, EM Santos, K Fukui. “Tensor Fukunaga-Koontz Transform for Hierarchical Clustering”. BRACIS 2019 (Brazilian Conference on Intelligent Systems)

Bioacoustic signals representation and classification: Here, bioacoustic signals classification methods based on subspaces are developed, where no pre-processing techniques are required. I attempt to demonstrate that this design achieves a compact representation for such signals, which is independent of the signal length. Besides, segmentation, noise reduction techniques, nor syllable extraction are needed. This project is mainly developed with prof. Juan G. Colonna (https://scholar.google.com/citations?user=wQdRmk4AAAAJ), who is responsible for data collection and curation (frogs and insects sound). We show that this method is theoretically and practically attractive through experimental results. The following papers present our primary findings:

 [7] BB Gatto, EM Santos, JG Colonna, N Sogi, LS Souza, K Fukui. “Discriminative Singular Spectrum Analysis for Bioacoustic Signal Classification”. INTERSPEECH 2020 (International Speech Communication Association)

[12] LS Souza, BB Gatto, K Fukui. “Classification of Bioacoustic Signals with Tangent Singular Spectrum Analysis”. ICASSP 2019 (International Conference on Acoustics, Speech, and Signal Processing)

[15] LS Souza, BB Gatto, K Fukui. “Grassmann singular spectrum analysis for bioacoustics classification”. ICASSP 2018 (International Conference on Acoustics, Speech, and Signal Processing)

[21] BB Gatto, JG Colonna, EM dos Santos, EF Nakamura. “Mutual singular spectrum analysis for bioacoustics classification”. MLSP 2017 (International Workshop on Machine Learning for Signal Processing)

Shallow networks based on subspaces: I develop a relationship between subspace-based methods and CNNs in image recognition tasks. More precisely, I demonstrate that we can efficiently employ subspace-based methods to construct CNNs, decreasing the amount of time required for training. This approach drastically reduces the learning parameters necessary to use a CNNs. The following papers present my findings:

 [1] BB Gatto, EM Santos, K Fukui, WSS Júnior, KV Santos. “Fukunaga-Koontz convolutional network with applications on character classification”. NEPL 2020 (Neural Processing Letters, Springer)

[2] BB Gatto, LS Souza, EM Santos, K Fukui, WSS Júnior, KV Santos. “A semi-supervised convolutional neural network based on subspace representation for image classification”. JIVP 2020 (EURASIP Journal on Image and Video Processing, Springer)

[20] BB Gatto, EM dos Santos, K Fukui. “Subspace-based convolutional network for handwritten character recognition”. ICDAR 2017 (International Conference on Document Analysis and Recognition)

[23] BB Gatto, EM Santos. “Discriminative Canonical Correlation Analysis Network for Image Classification”. ICIP 2017 (International Conference on Image Processing)

[24] BB Gatto, LS Souza, EM Santos. “A Deep Network Model based on Subspaces: A Novel Approach for Image Classification”. MVA 2017 (International Conference on Machine Vision Applications)

1. JSPS KAKENHI (Grants-in-Aid for Scientific Research) Awarded by the Japanese Government (2019-2023)

2. CAPES (Coordination for the Improvement of Higher Education) Awarded by the Brazilian Government to research in Japan (2017-2018)

3. FAPEAM (Foundation for Research Support of the State of Amazonas) Awarded by the Brazilian Government (2014-2018)

4. MEXT Scholarship Awarded by the Japanese Government (2011-2013)

Journal papers

[1]  BB Gatto, JG Colonna, EM dos Santos, AL Koerich, K Fukui. "Discriminative Singular Spectrum Classifier with applications on bioacoustic signal recognition". DSP 2023 (Digital Signal Processing)  (link)

[2]  LS Souza, N Sogi, BB Gatto, T Kobayashi, K Fukui. "Grassmannian learning mutual subspace method for image set recognition".  Neurocomputing 2023  (link)

[1] BB Gatto, EM Santos, K Fukui, WSS Júnior, KV Santos. “Fukunaga-Koontz convolutional network with applications on character classification”. NEPL 2020 (Neural Processing Letters, Springer) (link)

[2] BB Gatto, LS Souza, EM Santos, K Fukui, WSS Júnior, KV Santos. “A semi-supervised convolutional neural network based on subspace representation for image classification”. JIVP 2020 (EURASIP Journal on Image and Video Processing, Springer) (link)

[3] LS Souza, BB Gatto, JH Xue, K Fukui. “Enhanced Grassmann discriminant analysis with randomized time warping for motion recognition”. PR 2020 (Pattern Recognition, Elsevier) (link)

[4] MAF Mollinetti, MTRS Neto, BB Gatto, T Kuno. “A-DVM: A Self-adaptive Variable Matrix Decision Variable Selection Scheme for Multimodal Problems”. Entropy 2020 (Entropy, MDPI) (link)

[5] BB Gatto, EM Santos, MAF Molinetti, K Fukui. “Multilinear Clustering via Tensor Fukunaga-Koontz Transform with Fisher Eigenspectrum Regularization”. ASOC 2020 (Applied Soft Computing, Elsevier) status: under review.

[6] BB Gatto, EM Santos, AL Koerich, K Fukui, WSS Júnior. “Tensor Analysis with n-mode Generalized Difference Subspace”. ESWA 2020 (Expert Systems with Applications, Elsevier) status: under review. (preprint link)

International conferences

[7] BB Gatto, EM Santos, JG Colonna, N Sogi, LS Souza, K Fukui. “Discriminative Singular Spectrum Analysis for Bioacoustic Signal Classification”. INTERSPEECH 2020 (International Speech Communication Association) status: accepted.

[8] LS Souza, N Sogi, BB Gatto, K Fukui. “An Interface between Grassmann manifolds and vector spaces”. CVPRW 2020 (International Workshop on Differential Geometry in Computer Vision and Machine Learning) (link)

[9] N Sogi, LS Souza, BB Gatto, K Fukui “Metric Learning with A-based Scalar Product for Image-set Recognition”. CVPRW 2020 (International Workshop on Differential Geometry in Computer Vision and Machine Learning) (link)

[10] BB Gatto, MAF Molinetti, EM Santos, K Fukui. “Tensor Fukunaga-Koontz Transform for Hierarchical Clustering”. BRACIS 2019 (Brazilian Conference on Intelligent Systems) (link)

[11] MAF Mollinetti, MTRS Neto, BB Gatto, T Kuno. “Selecting Decision Variables for Artificial Bee Colony using a Self-adaptive Variable Matrix”. ENIAC 2019 (National Meeting on Artificial and Computational Intelligence) (link)

[12] LS Souza, BB Gatto, K Fukui. “Classification of Bioacoustic Signals with Tangent Singular Spectrum Analysis”. ICASSP 2019 (International Conference on Acoustics, Speech, and Signal Processing) (link)

[13] EK Shimomoto, LS Souza, BB Gatto, K Fukui. “News2meme: An Automatic Content Generator from News Based on Word Subspaces from Text and Image”. MVA 2019 (International Conference on Machine Vision Applications) (link)

[14] EK Shimomoto, LS Souza, BB Gatto, K Fukui. “Text classification based on word subspace with term-frequency". IJCNN 2018 (International Joint Conference on Neural Networks) (link)

[15] LS Souza, BB Gatto, K Fukui. “Grassmann singular spectrum analysis for bioacoustics classification”. ICASSP 2018 (International Conference on Acoustics, Speech, and Signal Processing) (link)

[16] IAS Filho, BB Gatto, JLS Pio, EN Chen, JM Júnior. “Gesture recognition using leap motion: a machine learning-based controller interface”. SETIT 2018 (Sciences of Electronics, Technologies of Information and Telecommunications) (link)

[17] IAS Filho, BB Gatto, JLS Pio, EN Chen, JM. Júnior. “A Interfacing Controller for Games based on Gesture Recognition”. IADIS 2018 (International Conference Applied Computing) (link)

[18] BB Gatto, EM Santos, WSS Júnior. “Orthogonal hankel subspaces for applications in gesture recognition”. SIBGRAPI 2017 (Conference on Graphics, Patterns and Images) (link)

[19] MM Yvano, BB Gatto, EM Santos. “A Method to Detect Boats in Images of the Amazonian Rivers”. BRACIS 2017 (Brazilian Conference on Intelligent Systems) (link)

[20] BB Gatto, EM dos Santos, K Fukui. “Subspace-based convolutional network for handwritten character recognition”. ICDAR 2017 (International Conference on Document Analysis and Recognition) (link)

[21] BB Gatto, JG Colonna, EM dos Santos, EF Nakamura. “Mutual singular spectrum analysis for bioacoustics classification”. MLSP 2017 (International Workshop on Machine Learning for Signal Processing) (link)

[22] BB Gatto, A Bogdanova, LS Souza, EM Santos. “Hankel subspace method for efficient gesture representation”. MLSP 2017 (International Workshop on Machine Learning for Signal Processing) (link)

[23] BB Gatto, EM Santos. “Discriminative Canonical Correlation Analysis Network for Image Classification”. ICIP 2017 (International Conference on Image Processing) (link)

[24] BB Gatto, LS Souza, EM Santos. “A Deep Network Model based on Subspaces: A Novel Approach for Image Classification”. MVA 2017 (International Conference on Machine Vision Applications) (link)

[25] LS Souza, BB Gatto, K Fukui. “Enhancing Discriminability of Randomized Time Warping for Motion Recognition”. MVA 2017 (International Conference on Machine Vision Applications) (link)

[26] BB Gatto, EM dos Santos. “Image-Set Matching by Two Dimensional Generalized Mutual Subspace Method”. BRACIS 2016 (Brazilian Conference on Intelligent Systems) (link)

[27] BB Gatto, WSS Júnior, EM Santos. “Kernel Two Dimensional Subspace for Image Set Classification”. ICTAI 2016 (International Conference on Tools with Artificial Intelligence) (link)

[28] JG Colonna, BB Gatto, EM Santos, EF Nakamura. “A framework for chainsaw detection using one-class kernel and wireless acoustic sensor networks into the amazon rainforest”. MDM 2016 (IEEE International Conference on Mobile Data Management) (link)

[29] JG Colonna, BB Gatto, EF Nakamura, EM dos Santos. “A framework for chainsaw detection using one-class and WSNs”. IPSN 2016 (International Conference on Information Processing in Sensor Networks) (link)

1. Best paper award: Bernardo B. Gatto, Marco A. F. Molinetti, Eulanda M. dos Santos, Kazuhiro Fukui, Tensor Fukunaga-Koontz Transform for Hierarchical Clustering, Brazilian Conference on Intelligent Systems (BRACIS 2019)

2. Best poster award: Erica K. Shimomoto, Lincon S. Souza, Bernardo B. Gatto, Kazuhiro Fukui, News2meme: An Automatic Content Generator from News Based on Word Subspaces from Text and Image IAPR MVA (MVA 2019)

3. Best poster award: Naoya Sogi, Lincon S. Souza, Bernardo B. Gatto, Rui Zhu, Jing-Hao Xue, Kazuhiro Fukui, 凸錐判別分析に基づく画像セットベース識別, Image-set Classification based on Convex Cone Discriminant Analysis, (MIRU 2019)

4. Student travel award: Discriminative Canonical Correlation Analysis Network for Image Classification (ICIP 2017, IEEE Signal Processing Society)

5. Best poster award: Bernardo B. Gatto, Lincon S. Souza, Eulanda M. dos Santos, A Deep Network Model based on Subspaces: A Novel Approach for Image Classification, IAPR MVA (MVA 2017)

6. Best poster award: Lincon S. Souza, Bernardo B. Gatto, Kazuhiro Fukui, Enhancing Discriminability of Randomized Time Warping for Motion Recognition, IAPR MVA (MVA 2017)

7. Best poster award: Bernardo B. Gatto, Eulanda M. dos Santos, A Deep Network Model based on Subspaces, United Kingdom (BMVA CVSS 2016)

8. Third place in the Latin American Doctoral Thesis Contest (https://clei2021.cr/CLTD)

9. Ph.D. Thesis Honorable Mention Award of the Workshop of Theses and Dissertations (WTD) (https://www.inf.ufrgs.br/sibgrapi2021/awards.php)