Archive

Posts Tagged ‘Research’

New paper accepted

December 14, 2012 1 comment

We got a paper accepted on the IEEE Trans. on Image Processing 🙂 . It is about a new micro pattern descriptor that relies on directional numbers (direction indexes in a local neighborhood), and we tested it on face analysis tasks, namely, face identification and expression recognition. We’ve been working on these ideas for some time, and sadly the revision was delayed some time, but finally it’s over.

I hope to give more details next week when I have more time, and give some insights on the ideas behind. Meanwhile, I leave you with the abstract and the pre-print at our site or at the IEEEXplore

Abstract—This paper proposes a novel local feature descriptor, Local Directional Number Pattern (LDN), for face analysis: face and expression recognition. LDN encodes the directional information of the face’s textures (i.e., the texture’s structure) in a compact way, producing a more discriminative code than current methods. We compute the structure of each micro-pattern with the aid of a compass mask, that extracts directional information, and we encode such information using the prominent direction indexes (directional numbers) and sign—which allows us to distinguish among similar structural patterns that have different intensity transitions. We divide the face into several regions, and extract the distribution of the LDN features from them. Then, we concatenate these features into a feature vector, and we use it as a face descriptor. We perform several experiments in which our descriptor performs consistently under illumination, noise, expression, and time lapse variations. Moreover, we test our descriptor with different masks to analyze its performance in different face analysis tasks.

Advertisements

Statistical Map: for edge-segment background modeling

November 19, 2012 Leave a comment

The first problem I tackled since I came to the lab was background modeling through edges. I worked on this problem for a couple of years, and a few papers on the road, I think it’s time to talk a little about it. It was a fun experience, being new to research and all that. However, it was a tough one as it is not a “hot topic” or even approved by the community who works on an opposite direction. Although, there are many possibilities by working with this type of information, we need further research to make it viable for real time applications.

Edge-based modeling

In the literature, the most common methods are pixel-based. That means, that the information used to model the scene is that contained in each pixel. Most methods use the intensity and its changes to create a model of the entire scene.

On the other hand, we are exploring (or should I say explored…) the use of edges as a reliable source of modeling. As the edges reveal the structure of the scene, we can model it with less information and less bulk. However, due to the lack of other non-structural information (pixel intensities) it may be harder to differentiate some objects with similar structure (about this later). Moreover, because we are using less pixels (as the edges in a scene are sparse) we can use more expressive models to represent it.

However, this approach is not mainstream, and is it hard to produce models on it, because you need to pre-process the images to obtain the edges, and then post-process it to obtain the regions. But if you are not interested in the regions and only need higher information, e.g., behavior information, the isolated structures are goodies that you can’t refuse.

Statistical Map

Overall flow

Fig. 1. Overall flow

The main idea behind the edge modeling is the use of a frequency map of the appearance of the edges, that we called Statistical Map (SM). The SM models the edge behavior over time. Ideally, it would create wide distributions for high movement edges, and narrow ones for edges with no movement. From the temporal point of view, the distributions will be high if the edge belongs to background and small if the edge is a moving one. Fig. 1 illustrates this idea.

The use of the SM is the basic model which can be extended by using color, intensity, gradient and other information, as the main problem with edges happens when moving edges get close to the borders of background. Thus, we explore most of these options to enhance the detection, as objects with similar structure will present different color (or other properties).

A more robust model

We increase the robustness of the model by adding memory to the model. We keep a set of lists of the edges that we saw before, as well as their shape, color, gradient, and other information, to build a more discriminative model. Therefore, we can use these lists to match and learn new edges as they appear in the sequence.

Publications

For more information about these methods you can refer to my papers:

Ramirez Rivera, A.; Murhsed, M.; Kim, J.; Chae, O., “Background Modeling through Statistical Edge-Segment Distributions,” Transactions on Circuits and Systems for Video Technology, 2012 (to appear). Preprint link

Ramirez Rivera, A.; Murshed, M.; Chae, O., “Object detection through edge behavior modeling,” AVSS, 2011. Paper link

or for more information check the other related papers at here.

For the BibTeX form, you can use this:

@ARTICLE{RamirezR2012,
author={Ramirez Rivera, A. and Murshed, M. and Kim, J. and Chae, O.},
journal={Circuits and Systems for Video Technology, IEEE Transactions on},
title={Background Modeling through Statistical Edge-Segment Distributions},
year={2012}
}

@INPROCEEDINGS{RamirezR2011,
title={Object detection through edge behavior modeling},
author={Ramirez Rivera, A. and Murshed, M. and Chae, O.},
booktitle={Advanced Video and Signal-Based Surveillance ({AVSS}), 2011 8th IEEE International Conference on},
pages={273--278},
year={2011},
organization={IEEE}
}

Birth of a word

March 11, 2011 Leave a comment

When several disciplines are merged together, great things could happen. It is not only the context and frequencies of word repetition, for a new word to be learned by a kid, but all the things that can be unveiled from this work. A lot of anthropologist  and psychologist will have a lot of to play with, if this technology can be scalable.

I just glad to see what people can do just with speech recognition, image analysis, and semantic analysis. You can mine and learn, not just words, but behavior. We can learn things about ourselves that we don’t even suspect.

I think that the limit is our imagination, on how we mix everything that is out there. The problem will be how to reach all that knowledge and being able to process it in a coherent way.

Image Enhancement Survey

February 23, 2011 3 comments

Hi, I’m doing a survey to assess a new enhancement algorithm for dark images that I’ve been working on.

So, please take a couple of minutes and help me. It contains 10 group of images, and you need to pick the one you prefer the most among them, in four different categories.

Thanks… :mrgreen:

Background Subtraction

October 5, 2010 Leave a comment

I just found a great compilation of Background Subtraction papers.

  • Background Models
    • Basic
    • Statistical
    • Fuzzy
    • Neural Network
    • Background Modeling via Clustering
    • Background Estimation
  • Features
    • Color
    • Edge
    • Textures
    • Motion
    • Stereo
  • Evaluation Techniques
  • Surveys

It has recent papers. It is worth to look at it. 💡

What is a Ph.D. about?

September 6, 2010 Leave a comment

I’ve been thinking what a Ph.D. is about. A couple of months ago, I read something regarding graduate studies and what people expect from them and what they really are. Yesterday, I read, again, a post of how to fail a Ph.D., and those readings come back to me. From all these sources I realize that what you will learn or the contributions you will make don’t matter, but what matters is the training. A Ph.D. prepares you to conduct quality research and gives you the tools to do it by yourself: to identify valuable problems, to solve those problems and to survive peer review.

Looking back, I can say that is similar to undergrad studies. In college, at least in my country and my former university, professors try to teach you how to learn by yourself and being able to overcome problems. If possible, they try to teach you some “new” ideas. The problem is that, after 5 years of college, you end up with a lot of useless knowledge because the technology and topics you learnt, probably, shifted to new ones; needing you to learn them again. The useful tools, however, that you learnt back then will paid now. Thanks to those tools you are able to adapt to changes and overcome new problems. I think that’s one of the reasons why some people succeed in (dynamic environment) works that demand new knowledge and why some others are stuck in crappy jobs doing the same thing; because they are not able to learn the new things by themselves.

In the end, all these years of school and education are not about the “new” topics you are learning, but building a set of tools to explore and try to advance forward. And if you have enough courage, make the humanity knowledge advance with you. I remember my advisor words: “…at the end of their Ph.D. most people work in something totally different of what they did…“, so it is not about the topic, but the skills you are building. 💡

%d bloggers like this: