Report on the International Conference on Pattern Recognition Systems (ICPRS)
11th - 13th July 2017
Report written by Marcos González Montero
The origin of the conference
First of all, to talk about the International Conference on Pattern Recognition Systems
(ICPRS) it’s necessary to go back to it origins. The Chilean Computer Science Society has
for many years organised an annual conference and in each conference there were different
workshops reflecting specialised fields.
Some years later, and thanks to the importance that its workshop gained in a short
period of time and to the amount of papers that came from many countries to its workshop,
the Chilean Association for Pattern Recognition (ACHIRP being its acronym in Spanish)
emerged as an independent organization and associated to the International Association
for Pattern Recognition (IAPR). As of 2016, the workshop became a separate conference.
The IET’s Vision and Imaging Network has supported ACHIRP’s workshops and then its
conference through publicity and publication of accepted papers.
The ICPRS has its headquarters in Chile for two basic reasons: on the one hand to
maintain the local character of its origins, on the other hand because of the high cost of the
flights and trips between Latin America and United States or the European Union. This
second point is well known by the conference attendants, as I could check.
One of the conference organizers, Dr Marcelo Mendoza who is also ACHIRP’s president,
told me that there are three basic points to understand the evolution and the improvement
of the conference: there is a very nice organization behind the scenes that enables the success
in each event, based on collaborative work, so this kind of collaboration makes the conference
grows; the improvement is due to the internal links in the ACHIRP and the essential IET
collaboration too.
Thanks to all these things -collaborations and organization- is possible to publish the
papers and posters in the IEEE and IET repositories at the same time than in the ICPRS.
This enables to give more visibility and prestige to the conference than otherwise it would
not have been possible. As a result, ICPRS-16 and ICPRS-17 received many papers from
different parts of the world.
To emphasize the idea of the collaboration, it’s important to mention that the 8th ICPRS
took place outside Chile for the first time and held in Madrid between 11-13th July 2017.
Sensations during the conference
From my point of view, someone who had finished the bachelor degree one week before the
conference, and from the point of view of a person who hasn’t got a technical knowledge
about the terms and themes of the conference I think that the main feeling can be very
similar, even the same: the good sensation about the relationships between the attendees.
In the conference there were people from many countries like India, China, France, Chile,
Spain, UK, Puerto Rico, USA, Canada, Algeia etc. Each person belongs to a research
group at University and it’s important to mention that this kind of conferences are directly
orientated to the academic investigation instead of industrial research or business production,
although the researches reach the real world through business. It is very interesting to see
how people interact between them during the coffee breaks and lunches talking about the
posters and papers, explaining each other their projects deeply and talking about things other
than work such as the possibility of having something to eat or drink after the meeting. I
liked very much how people enjoy this kind of meetings and how they can learn from each
other’s researches.
The lines of investigation are very different, as their applications, but all of them have a
base point in common: pattern recognition, as the name of the conference indicates. Almost
all of the studies that were presented and the presentations I saw proposed to solve real
and important problems like the automatic classification of seismic signals, human action
recognition in realistic scenarios and pedestrian detection inter alia.
Finally, and from a more technical point of view, I’ve to remark the variability of methods
that were presented. I mean, the researchers are using for example classic classifiers like
support vector machines or random forest (this kind of methods are robust and have very
good results in the literature) and moreover some artificial neural networks (a currently
famous method that has very good results and a high versatility); even in some cases we can
see a collaboration between the old and new classification methods improving the results in
many studies.
Summarizing, I’m convinced that the conference is the result of a very good and a very
hard work by the ACHIP and the IET and the local organiser that brought success. Much
more considering that this was the first time that the conference was held outside Chile and
the attendants were very satisfied with the results, the conference rooms and with the staff
work. So I have to thank the organizers and the IET for the great job that they have done
and the opportunity they give people like me to interact with experts in the field. I would
also like to invite people to go to ICPRS-18 in the picturesque city of Valparaiso in Chile on
May next year.