Home

Hi! I'm machine learning researcher.

This site contains information regarding my research. It contains code, data sets and my publications.

Here's a subset of my research interests:
- Multiple instance learning
- Action recognition
- Reinforcement learning
- Feature learning
- Computer vision

I'm affiliated with two labs:
  • LIVIA -  Laboratory for Imagery, Vision and Artificial Intelligence [link]
  • LACIME - Communication and Microelectronic Integration Laboratory [link]
Here is my google scholar profile.
Here is my Linkedin profile.


News
10/11/2017: We posted a paper combining active learning and multiple instance learning on ArXiv.

10/11/2017: Our paper Feature learning from spectrograms for assessment of personality traits got accepted for publication in IEEE Transactions on Affective Computing.

10/10/2017: Our survey paper on multiple instance learning got accepted for publication in Elsevier's Pattern Recognition.

9/11/2017: I just defended my thesis! It went well and I got the mention excellent!

6/8/2017: I published the code used for the survey paper: Multiple Instance Learning: A Survey of Problem Characteristics and Applications. You can download it here.

2/10/2017:
I'm part of the organizing committee of LABELS. It is a workshop on large-scale annotation of biomedical data and expert label synthesis. The workshop is held in conjunction with MICCAI 2017.

1/22/2016: I've been nominated publication chairman for the International Conference on Image Processing Theory, Tools and Applications (IPTA'17), which will be held in Montreal on November 28th to December 1st. (http://www.ipta-conference.com/ipta17/)

1/19/2017: We've put two papers on arXiv. The first one is a survey on multiple instance learning describing the types of problems and challenges associated with it. We also review methods addressing the challenges and how these challenges manifest in different application domains.
Here's the link to the paper: Multiple Instance Learning: A Survey of Problem Characteristics and Applications.

The second paper presents a methods for feature learning in personality assessment from speech based on dictionary learning and sparse coding. It can be found at: Feature learning from spectrograms for assessment of personality traits.

7/15/2016: My paper "Witness Identification in Multiple Instance Learning Using Random Subspaces" have been accepted at ICPR 2016, which will be held at Cancun, Mexico on December 4-8.

6/1/2016: The paper "Robust multiple-instance learning ensembles using random subspace instance selection" published in Pattern Recognition has been selected to be presented at the 2nd Global Summit and Expo on Multimedia & Applications held August 15-16, 2016 in London, UK.

5/20/2016: I'll be chairman for a special session called "Multiple Instance Learning for Image Analysis" at the sixth International Conference on Image Processing Theory, Tools and Applications (IPTA'16) held on December 12-15, 2016, at Oulu (Finland). I encourage fellow MIL researchers to submit papers and come to meet other MIL enthusiasts