smArDS is an advanced audio-visual tool that provides the best time and place to insert publicity in entertainment videos.
A software specially designed for increasing the number of insertion points in video streaming services while maintaining your Video Ad Completion Rates.
Extracting football highlights to generate automatic summaries
Video football sequences to be broadcasted in TV sports news, are analyzed and summarized manually which implies a laborious and exhausting task. We have developed an automatic summarization system of football highlights using audio-visual descriptors which helps to automate the summarization process.
Analyzing microchip layers to understand their layout
We have developed a software for automatic reconstruction of a multilayer circuit from thousands of partial microscopic images. The system reconstructs the whole layers of the integrated circuit (IC) identifying logical gates and its interconnections to obtain the final schematic required in reverse engineering processes. The output is given in GDS and Verilog format and is readable by commercial applications such as Cadence.
2012 - 2014
Face Recognition in TV
Monitoring the appearances of politics in TV
Politics is an Ugiat Technologies website that presents the percentages of time devoted to different politicians by the main television channels news programs. The great novelty of the system is that the statistics are fully automatic, they are extracted without any human intervention but using a proprietary face recognition system based on Deep Learning.
Decomposing TV Shows in scenes, shots and keyframes and describing its contents
Videos are hierarchically decomposed in comprehensible temporal units: scenes, shots and keyframes. Then, each unit is summarized by a set of valuable metadata. For instance, the most important actors are facially recognized, global locations where the action takes place are described, objects found in the image are detected, perceived quality of the photography is scored and more.
Legal and Privacy
This web site is managed by Ugiat Technologies that can be contacted through the following mail: firstname.lastname@example.org
The content of this web site are mainly images, text, logos, graphics and other computer applications. The right of these media files as well as the domain name belong to the web site owner aside from some images obtained through public images databases such as www.freeimages.com.
This website is developed using Bootstrap that works under the MIT License.
Personal Data Privacy
This website does not store any personal information execept some implicit data. It could be the IP navigation address required to transmit the data for navigating through the web as well as the information gathered by external applications as Google Analytics which is out of our supervision.
The web site owner do not assure the reliability, availability and continuity of the web site contents arose from technical issues, security problems, hacking attacks, hosting problems and other relevant causes out of the knowledge of the web site owner.