<?xml version="1.0" encoding="UTF-8"?>
<feed xmlns:yt="http://www.youtube.com/xml/schemas/2015" xmlns:media="http://search.yahoo.com/mrss/" xmlns="http://www.w3.org/2005/Atom">
 <link rel="self" href="http://www.youtube.com/feeds/videos.xml?channel_id=UCzd1PDbjInGwU5mLLea_Cyw"/>
 <id>yt:channel:UCzd1PDbjInGwU5mLLea_Cyw</id>
 <yt:channelId>UCzd1PDbjInGwU5mLLea_Cyw</yt:channelId>
 <title>Vicomtech-IK4</title>
 <link rel="alternate" href="https://www.youtube.com/channel/UCzd1PDbjInGwU5mLLea_Cyw"/>
 <author>
  <name>Vicomtech-IK4</name>
  <uri>https://www.youtube.com/channel/UCzd1PDbjInGwU5mLLea_Cyw</uri>
 </author>
 <published>2008-06-06T10:43:34+00:00</published>
 <entry>
  <id>yt:video:RjludQUSI9U</id>
  <yt:videoId>RjludQUSI9U</yt:videoId>
  <yt:channelId>UCzd1PDbjInGwU5mLLea_Cyw</yt:channelId>
  <title>SSABi - Servicio de Subtitulación Automática Bilingüe</title>
  <link rel="alternate" href="https://www.youtube.com/watch?v=RjludQUSI9U"/>
  <author>
   <name>Vicomtech-IK4</name>
   <uri>https://www.youtube.com/channel/UCzd1PDbjInGwU5mLLea_Cyw</uri>
  </author>
  <published>2017-07-21T08:10:03+00:00</published>
  <updated>2017-08-05T14:10:02+00:00</updated>
  <media:group>
   <media:title>SSABi - Servicio de Subtitulación Automática Bilingüe</media:title>
   <media:content url="https://www.youtube.com/v/RjludQUSI9U?version=3" type="application/x-shockwave-flash" width="640" height="390"/>
   <media:thumbnail url="https://i3.ytimg.com/vi/RjludQUSI9U/hqdefault.jpg" width="480" height="360"/>
   <media:description>SSABi es una plataforma para la subtitulación automática de contenidos bilingües, en euskera y castellano. Está basada íntegramente en técnicas de procesamiento del habla, incluyendo tecnología de reconocimiento del habla, detección de cambio de idioma, puntuación y capitalización, y generación de subtítulos que cumplen con los estándares de calidad exigidos. SSABi reduce drásticamente el tiempo y coste necesarios para la generación de subtítulos de contenidos audiovisuales frente al método tradicional de subtitulación manual.

Más información:
language@vicomtech.org

Transkit® Demo:
http://www.speechandlanguagesolutions.com/automatic-transcription.html

Transkit® SDK brochure:
http://www.vicomtech.org/resources/archivosbd/sdks_documentos/transkit_sdk.pdf</media:description>
   <media:community>
    <media:starRating count="0" average="0.00" min="1" max="5"/>
    <media:statistics views="68"/>
   </media:community>
  </media:group>
 </entry>
 <entry>
  <id>yt:video:MRZxoK5ZjPs</id>
  <yt:videoId>MRZxoK5ZjPs</yt:videoId>
  <yt:channelId>UCzd1PDbjInGwU5mLLea_Cyw</yt:channelId>
  <title>Solución Inteligente de Atención Post-Venta al Cliente Internacional</title>
  <link rel="alternate" href="https://www.youtube.com/watch?v=MRZxoK5ZjPs"/>
  <author>
   <name>Vicomtech-IK4</name>
   <uri>https://www.youtube.com/channel/UCzd1PDbjInGwU5mLLea_Cyw</uri>
  </author>
  <published>2017-07-19T09:15:21+00:00</published>
  <updated>2017-07-19T10:30:21+00:00</updated>
  <media:group>
   <media:title>Solución Inteligente de Atención Post-Venta al Cliente Internacional</media:title>
   <media:content url="https://www.youtube.com/v/MRZxoK5ZjPs?version=3" type="application/x-shockwave-flash" width="640" height="390"/>
   <media:thumbnail url="https://i2.ytimg.com/vi/MRZxoK5ZjPs/hqdefault.jpg" width="480" height="360"/>
   <media:description>IRTEN es una solución modular que permite a las PYMES optimizar sus servicios post-venta de atención al cliente internacional, disminuyendo el número de consultas que requieren de operadores especializados, facilitando su labor y reduciendo los costes asociados. Se basa en herramientas y tecnología para facilitar los procesos de customización, atención automática, atención manual, seguimiento y traducción del servicio de atención.

Más información:
language@vicomtech.org</media:description>
   <media:community>
    <media:starRating count="0" average="0.00" min="1" max="5"/>
    <media:statistics views="84"/>
   </media:community>
  </media:group>
 </entry>
 <entry>
  <id>yt:video:6BXkhykKMUw</id>
  <yt:videoId>6BXkhykKMUw</yt:videoId>
  <yt:channelId>UCzd1PDbjInGwU5mLLea_Cyw</yt:channelId>
  <title>New generation, low-cost, lane-level, precise turn-by-turn navigation application</title>
  <link rel="alternate" href="https://www.youtube.com/watch?v=6BXkhykKMUw"/>
  <author>
   <name>Vicomtech-IK4</name>
   <uri>https://www.youtube.com/channel/UCzd1PDbjInGwU5mLLea_Cyw</uri>
  </author>
  <published>2017-07-17T11:51:46+00:00</published>
  <updated>2017-07-28T16:42:18+00:00</updated>
  <media:group>
   <media:title>New generation, low-cost, lane-level, precise turn-by-turn navigation application</media:title>
   <media:content url="https://www.youtube.com/v/6BXkhykKMUw?version=3" type="application/x-shockwave-flash" width="640" height="390"/>
   <media:thumbnail url="https://i3.ytimg.com/vi/6BXkhykKMUw/hqdefault.jpg" width="480" height="360"/>
   <media:description>Development of a new generation, low-cost, lane-level, precise turn-by-turn nvigation application through the fusion of EGNSS and Computer Vision technology. This will enable a new generation of enhanced mapping information with real-time updating based on crowdsourcing techniques. The resulting lane-level vehicle positioning will bring navigation and traffic management systems to a new level of detail and effectiveness.

inLane targets at reaching the lane level accuracy, thanks to the use and development of complex fusion an hybridisation algorithms for GNSS, IMU, Map and Computer Vision signal.

More info:
http://inlane.eu
info@inlane.eu

Consortium
Vicomtech
Tomtom
Ertico
Intel
HRI
Ifsttar
Ajuntament de Barcelona
TU/e
RACC
Teleconsult

More info:
Viulib® is a software library, a solution that collects, processes and analyzes real-time video images.
Viulib®: http://www.viulib.org/
Viulib® SDK brochure: http://www.vicomtech.org/resources/archivosbd/sdks_documentos/viulib_sdk.pdf
Viulib® is Property of http://www.vicomtech.org
Contact for Viulib®: viulib@vicomtech.org</media:description>
   <media:community>
    <media:starRating count="3" average="5.00" min="1" max="5"/>
    <media:statistics views="72"/>
   </media:community>
  </media:group>
 </entry>
 <entry>
  <id>yt:video:LopmUqhEF7I</id>
  <yt:videoId>LopmUqhEF7I</yt:videoId>
  <yt:channelId>UCzd1PDbjInGwU5mLLea_Cyw</yt:channelId>
  <title>Resolving crimes with automatic person detection and re identification</title>
  <link rel="alternate" href="https://www.youtube.com/watch?v=LopmUqhEF7I"/>
  <author>
   <name>Vicomtech-IK4</name>
   <uri>https://www.youtube.com/channel/UCzd1PDbjInGwU5mLLea_Cyw</uri>
  </author>
  <published>2017-05-09T14:04:34+00:00</published>
  <updated>2017-07-11T10:43:56+00:00</updated>
  <media:group>
   <media:title>Resolving crimes with automatic person detection and re identification</media:title>
   <media:content url="https://www.youtube.com/v/LopmUqhEF7I?version=3" type="application/x-shockwave-flash" width="640" height="390"/>
   <media:thumbnail url="https://i1.ytimg.com/vi/LopmUqhEF7I/hqdefault.jpg" width="480" height="360"/>
   <media:description>This video presents a feature for management of crime reports, namely, the automatic correlation of crime reports based on person re-identification.

Two scenes are presented: a crime scene where car thieves were captured on a video taken by an accidental observer; and a second scene where the same actors are trying to sell a mobile on the street, resulting in a suspicious activity for an accidental citizen. Video and images from both scenes are sent to the police crime management center, currently implemented via the INSPEC2T platform. The person recognition feature of the INSPEC2T platform allows the police operators to get automatic notifications if the same person gets identified in images from different incident reports. As such, the operator gets automatic notification of related incidents or crimes and can proceed with the resolution of the related crime cases.  Besides the correlation of apparently not related incidents, the person re-identification feature may help to encounter missing persons or to reveal suspect criminals.

These methods are under development and continuously improving thanks to our participation in several EU and private projects: http://inspec2t-project.eu/


More info:
Viulib® is a software library, a solution that collects, processes and analyzes real-time video images.
Viulib®: http://www.viulib.org/
Viulib® SDK brochure: http://www.vicomtech.org/resources/archivosbd/sdks_documentos/viulib_sdk.pdf
Viulib® is Property of http://www.vicomtech.org
Contact for Viulib®: viulib@vicomtech.org
General contact: security@vicomtech.org</media:description>
   <media:community>
    <media:starRating count="2" average="5.00" min="1" max="5"/>
    <media:statistics views="397"/>
   </media:community>
  </media:group>
 </entry>
 <entry>
  <id>yt:video:xBUT4nJDh20</id>
  <yt:videoId>xBUT4nJDh20</yt:videoId>
  <yt:channelId>UCzd1PDbjInGwU5mLLea_Cyw</yt:channelId>
  <title>Pixelwise Annotator for Ground Truth Generation</title>
  <link rel="alternate" href="https://www.youtube.com/watch?v=xBUT4nJDh20"/>
  <author>
   <name>Vicomtech-IK4</name>
   <uri>https://www.youtube.com/channel/UCzd1PDbjInGwU5mLLea_Cyw</uri>
  </author>
  <published>2017-04-28T08:33:19+00:00</published>
  <updated>2017-05-25T20:02:50+00:00</updated>
  <media:group>
   <media:title>Pixelwise Annotator for Ground Truth Generation</media:title>
   <media:content url="https://www.youtube.com/v/xBUT4nJDh20?version=3" type="application/x-shockwave-flash" width="640" height="390"/>
   <media:thumbnail url="https://i1.ytimg.com/vi/xBUT4nJDh20/hqdefault.jpg" width="480" height="360"/>
   <media:description>Pixelwise annotation tool, using Viulib computer vision libraries (www.viulib.org).

This tool provides an easy-to-use graphical user interface for efficient annotation of images at pixel level. The use of superpixels algorithms allows to quickly define all the objects in the scene with high precision.

The tool contains a number of features to simplify and speed up the annotation process, such as possibility of choosing different superpixels algorithms and resolution, filling of predefined contours, and use of drawing layers.

Automatic segmentation of the image using deep learning technology is included. The user can ask for a pre-annotated result, which is computed using trained neural networks of high accuracy. Then, this result can be modified and refined by the user if neccessary. Annotation times are greatly reduced using this option.

More info:
Viulib® is a software library, a solution that collects, processes and analyzes real-time video images.
Viulib®: http://www.viulib.org/
Viulib® SDK brochure: http://www.vicomtech.org/resources/archivosbd/sdks_documentos/viulib_sdk.pdf
Viulib® is Property of http://www.vicomtech.org
Contact for Viulib®: viulib@vicomtech.org
General contact: its@vicomtech.org</media:description>
   <media:community>
    <media:starRating count="5" average="5.00" min="1" max="5"/>
    <media:statistics views="311"/>
   </media:community>
  </media:group>
 </entry>
 <entry>
  <id>yt:video:K4Zsd1w5dPk</id>
  <yt:videoId>K4Zsd1w5dPk</yt:videoId>
  <yt:channelId>UCzd1PDbjInGwU5mLLea_Cyw</yt:channelId>
  <title>Decision Support and Information Management System for Breast Cancer</title>
  <link rel="alternate" href="https://www.youtube.com/watch?v=K4Zsd1w5dPk"/>
  <author>
   <name>Vicomtech-IK4</name>
   <uri>https://www.youtube.com/channel/UCzd1PDbjInGwU5mLLea_Cyw</uri>
  </author>
  <published>2017-04-25T08:40:57+00:00</published>
  <updated>2017-04-25T11:16:25+00:00</updated>
  <media:group>
   <media:title>Decision Support and Information Management System for Breast Cancer</media:title>
   <media:content url="https://www.youtube.com/v/K4Zsd1w5dPk?version=3" type="application/x-shockwave-flash" width="640" height="390"/>
   <media:thumbnail url="https://i4.ytimg.com/vi/K4Zsd1w5dPk/hqdefault.jpg" width="480" height="360"/>
   <media:description>DESIREE project is an international initiative for breast cancer study and treatment improvement, which uses the latest available technology to help clinicians in their decision making process. This technology includes advanced clinical decision support systems (based on formalized clinical guidelines, knowledge discovery, and case based reasoning), image-based breast and tumour characterization tools and predictive modeling for 3D breast reconstruction. 


More info: http://www.desiree-project.eu/
    desire-contact@vicomtech.org

Partners:

Vicomtech-IK4 (Coordinator), Spain
Arivis AG, Germany
Hôpital Tenon, France
Bilbomatica, Spain
ERESA, Spain
Fundación Onkologikoa, Spain
INSERM-LIMICS, France
Medical Innovation and Technology, Greece
Sistemas Genómicos, Spain
University of Houston, USA
University of Ulster, UK</media:description>
   <media:community>
    <media:starRating count="1" average="5.00" min="1" max="5"/>
    <media:statistics views="107"/>
   </media:community>
  </media:group>
 </entry>
 <entry>
  <id>yt:video:uGvvCgqYmpY</id>
  <yt:videoId>uGvvCgqYmpY</yt:videoId>
  <yt:channelId>UCzd1PDbjInGwU5mLLea_Cyw</yt:channelId>
  <title>Deep Learning for Uncalibrated Gaze Estimation and Fatigue Detection</title>
  <link rel="alternate" href="https://www.youtube.com/watch?v=uGvvCgqYmpY"/>
  <author>
   <name>Vicomtech-IK4</name>
   <uri>https://www.youtube.com/channel/UCzd1PDbjInGwU5mLLea_Cyw</uri>
  </author>
  <published>2017-04-11T06:43:33+00:00</published>
  <updated>2017-05-05T15:52:37+00:00</updated>
  <media:group>
   <media:title>Deep Learning for Uncalibrated Gaze Estimation and Fatigue Detection</media:title>
   <media:content url="https://www.youtube.com/v/uGvvCgqYmpY?version=3" type="application/x-shockwave-flash" width="640" height="390"/>
   <media:thumbnail url="https://i2.ytimg.com/vi/uGvvCgqYmpY/hqdefault.jpg" width="480" height="360"/>
   <media:description>This video illustrates the work we are carrying out in the field of Driver Monitoring Systems (DMS) using single camera set-ups, and our latest advances in computer vision and machine learning. 

Driver attention is monitored by means of estimating the head pose and gaze estimation, for which we use deep learning methods for training inference models. The fatigue of the driver can be also measured processing the observed amplitude of the eyes, which provide information on blinking patterns we can correlate with fatigue levels.

These methods are under development and continuously improving thanks to our participation in several EU and private projects: 
http://vi-das.eu/

More info:
Viulib® is a software library, a solution that collects, processes and analyzes real-time video images.
Viulib®: http://www.viulib.org/
Viulib® SDK brochure: http://www.vicomtech.org/resources/archivosbd/sdks_documentos/viulib_sdk.pdf
Viulib® is Property of http://www.vicomtech.org
Contact for Viulib®: viulib@vicomtech.org
General contact: its@vicomtech.org</media:description>
   <media:community>
    <media:starRating count="16" average="5.00" min="1" max="5"/>
    <media:statistics views="1319"/>
   </media:community>
  </media:group>
 </entry>
 <entry>
  <id>yt:video:itXUsSJy7nA</id>
  <yt:videoId>itXUsSJy7nA</yt:videoId>
  <yt:channelId>UCzd1PDbjInGwU5mLLea_Cyw</yt:channelId>
  <title>Digital Twin of a robot and gripper</title>
  <link rel="alternate" href="https://www.youtube.com/watch?v=itXUsSJy7nA"/>
  <author>
   <name>Vicomtech-IK4</name>
   <uri>https://www.youtube.com/channel/UCzd1PDbjInGwU5mLLea_Cyw</uri>
  </author>
  <published>2017-03-13T08:43:15+00:00</published>
  <updated>2017-04-18T18:45:38+00:00</updated>
  <media:group>
   <media:title>Digital Twin of a robot and gripper</media:title>
   <media:content url="https://www.youtube.com/v/itXUsSJy7nA?version=3" type="application/x-shockwave-flash" width="640" height="390"/>
   <media:thumbnail url="https://i2.ytimg.com/vi/itXUsSJy7nA/hqdefault.jpg" width="480" height="360"/>
   <media:description>This demo shows a collaborative robot synchronized with its virtual representation. The real-time visualization is running on a Web browser which uses WebGL technology for 3D rendering and receives continous data about the state and pose of the robot and gripper. Through the web application a user can also send motion orders to the robot joints.

more info:
industry@vicomtech.org</media:description>
   <media:community>
    <media:starRating count="6" average="5.00" min="1" max="5"/>
    <media:statistics views="336"/>
   </media:community>
  </media:group>
 </entry>
 <entry>
  <id>yt:video:JJZQTGMU714</id>
  <yt:videoId>JJZQTGMU714</yt:videoId>
  <yt:channelId>UCzd1PDbjInGwU5mLLea_Cyw</yt:channelId>
  <title>Thematic mapping by large scale geospatial machine learning</title>
  <link rel="alternate" href="https://www.youtube.com/watch?v=JJZQTGMU714"/>
  <author>
   <name>Vicomtech-IK4</name>
   <uri>https://www.youtube.com/channel/UCzd1PDbjInGwU5mLLea_Cyw</uri>
  </author>
  <published>2017-02-28T12:54:36+00:00</published>
  <updated>2017-06-12T15:05:10+00:00</updated>
  <media:group>
   <media:title>Thematic mapping by large scale geospatial machine learning</media:title>
   <media:content url="https://www.youtube.com/v/JJZQTGMU714?version=3" type="application/x-shockwave-flash" width="640" height="390"/>
   <media:thumbnail url="https://i3.ytimg.com/vi/JJZQTGMU714/hqdefault.jpg" width="480" height="360"/>
   <media:description>This video shows an supervised real-time thematic mapping prototype based on user selection patterns via web interface. This requires the integration of machine learning capabilities within web map servers, along with a simple interface that enables navigation and the monitoring of geospatial learning. Semantic tagging algorithms based on supervised classification methods can be exploited for thematic map creation from raster data based on user needs.</media:description>
   <media:community>
    <media:starRating count="3" average="5.00" min="1" max="5"/>
    <media:statistics views="205"/>
   </media:community>
  </media:group>
 </entry>
 <entry>
  <id>yt:video:9FC0PN8n8sY</id>
  <yt:videoId>9FC0PN8n8sY</yt:videoId>
  <yt:channelId>UCzd1PDbjInGwU5mLLea_Cyw</yt:channelId>
  <title>Anomaly Event Detection using Spatio-Temporal Motion Patterns</title>
  <link rel="alternate" href="https://www.youtube.com/watch?v=9FC0PN8n8sY"/>
  <author>
   <name>Vicomtech-IK4</name>
   <uri>https://www.youtube.com/channel/UCzd1PDbjInGwU5mLLea_Cyw</uri>
  </author>
  <published>2017-01-30T12:44:52+00:00</published>
  <updated>2017-05-04T06:58:33+00:00</updated>
  <media:group>
   <media:title>Anomaly Event Detection using Spatio-Temporal Motion Patterns</media:title>
   <media:content url="https://www.youtube.com/v/9FC0PN8n8sY?version=3" type="application/x-shockwave-flash" width="640" height="390"/>
   <media:thumbnail url="https://i2.ytimg.com/vi/9FC0PN8n8sY/hqdefault.jpg" width="480" height="360"/>
   <media:description>The importance of anomaly event detection is growing up in recent years due to the improvement of security and safety in video surveillance systems. We focus on the assumption that the exhibition frequency is the essential characteristic of an event to be normal or abnormal. This application proposes an on-line and real-time normality modelling algorithm to deal with the anomaly detection problem in complex and crowded environments. Actions or events are interpreted as a set of spatio-temporal motion patterns in a neighbourhood, where a bag of words (BOV) approach is used to define a dictionary of common motion patterns. Them, these actions are modelled using a probabilistic framework in order to obtain the probability of being an abnormal behaviour. This anomaly probability is based on a Gaussian mixture model (GMM) which is able to add new behaviours appearing in the environment.

More info:
Viulib® is a software library, a solution that collects, processes and analyzes real-time video images. Download Viulib: http://www.viulib.org/
Viulib is Property of http://www.vicomtech.org
Contact for Viulib: viulib@vicomtech.org
General contact: its@vicomtech.org</media:description>
   <media:community>
    <media:starRating count="1" average="5.00" min="1" max="5"/>
    <media:statistics views="371"/>
   </media:community>
  </media:group>
 </entry>
 <entry>
  <id>yt:video:URdUbMK-R00</id>
  <yt:videoId>URdUbMK-R00</yt:videoId>
  <yt:channelId>UCzd1PDbjInGwU5mLLea_Cyw</yt:channelId>
  <title>Radiology on the move - Radmove project results</title>
  <link rel="alternate" href="https://www.youtube.com/watch?v=URdUbMK-R00"/>
  <author>
   <name>Vicomtech-IK4</name>
   <uri>https://www.youtube.com/channel/UCzd1PDbjInGwU5mLLea_Cyw</uri>
  </author>
  <published>2016-12-22T15:51:06+00:00</published>
  <updated>2017-02-27T20:23:24+00:00</updated>
  <media:group>
   <media:title>Radiology on the move - Radmove project results</media:title>
   <media:content url="https://www.youtube.com/v/URdUbMK-R00?version=3" type="application/x-shockwave-flash" width="640" height="390"/>
   <media:thumbnail url="https://i2.ytimg.com/vi/URdUbMK-R00/hqdefault.jpg" width="480" height="360"/>
   <media:description>RADMOVE project demonstrates the usability of current web applications technologies in order to facilitate &quot;Radiology on the move&quot; approaches to file exchange and radiological reporting in nowadays clinical practice. This development versión demonstration video illustrates the different functionalities: DICOM files upload&amp;sharing&amp;download, web anonymization, 2D and 3D visualization, medical annotations and collaborative reporting.</media:description>
   <media:community>
    <media:starRating count="4" average="5.00" min="1" max="5"/>
    <media:statistics views="207"/>
   </media:community>
  </media:group>
 </entry>
 <entry>
  <id>yt:video:AhIVW_2kINw</id>
  <yt:videoId>AhIVW_2kINw</yt:videoId>
  <yt:channelId>UCzd1PDbjInGwU5mLLea_Cyw</yt:channelId>
  <title>Visual Analytics for Data Value Envisioning</title>
  <link rel="alternate" href="https://www.youtube.com/watch?v=AhIVW_2kINw"/>
  <author>
   <name>Vicomtech-IK4</name>
   <uri>https://www.youtube.com/channel/UCzd1PDbjInGwU5mLLea_Cyw</uri>
  </author>
  <published>2016-09-08T08:32:51+00:00</published>
  <updated>2016-09-08T09:55:04+00:00</updated>
  <media:group>
   <media:title>Visual Analytics for Data Value Envisioning</media:title>
   <media:content url="https://www.youtube.com/v/AhIVW_2kINw?version=3" type="application/x-shockwave-flash" width="640" height="390"/>
   <media:thumbnail url="https://i2.ytimg.com/vi/AhIVW_2kINw/hqdefault.jpg" width="480" height="360"/>
   <media:description>Visual analysis of data from sensors and machines requires tools that can input multiple standard data formats, and output visual representations. The tools being developed at Vicomtech automatically discover relations across many variables and though time, group observed events/phenomena in data by similarity, and include forecasting and alarm generation abilities.</media:description>
   <media:community>
    <media:starRating count="0" average="0.00" min="1" max="5"/>
    <media:statistics views="197"/>
   </media:community>
  </media:group>
 </entry>
 <entry>
  <id>yt:video:PcxKRphxcw4</id>
  <yt:videoId>PcxKRphxcw4</yt:videoId>
  <yt:channelId>UCzd1PDbjInGwU5mLLea_Cyw</yt:channelId>
  <title>Real-time vehicle detection and lane detection for ADAS</title>
  <link rel="alternate" href="https://www.youtube.com/watch?v=PcxKRphxcw4"/>
  <author>
   <name>Vicomtech-IK4</name>
   <uri>https://www.youtube.com/channel/UCzd1PDbjInGwU5mLLea_Cyw</uri>
  </author>
  <published>2016-07-27T09:07:48+00:00</published>
  <updated>2017-07-21T01:10:14+00:00</updated>
  <media:group>
   <media:title>Real-time vehicle detection and lane detection for ADAS</media:title>
   <media:content url="https://www.youtube.com/v/PcxKRphxcw4?version=3" type="application/x-shockwave-flash" width="640" height="390"/>
   <media:thumbnail url="https://i1.ytimg.com/vi/PcxKRphxcw4/hqdefault.jpg" width="480" height="360"/>
   <media:description>Single camera detection of vehicles and lanes, using Viulib computer vision libraries (www.viulib.org).

This algorithm can be run into embedded systems, including Android and iOS devices. The system is able to detect the presence of multiple cars, track their position, and locate them into detected lanes. The distance to the vehicles is computed using a vanishing point-based calibration of the scene.

The lanes are modelled as parabolic curves in the bird's-eye view of the road. Ultra-fast linear algebra operations are used to minimize the computational load of the algorithm, which can run below 1 ms in most CPUs.

This method does not use special HW so it can be deployed easily on any platform or architecture and can be attached to ADAS systems that work with camera only sensors, or in combination with range sensors.

More info:
Viulib® is a software library, a solution that collects, processes and analyzes real-time video images. Download Viulib: http://www.viulib.org/
Viulib is Property of http://www.vicomtech.org
Contact for Viulib: viulib@vicomtech.org
General contact: its@vicomtech.org</media:description>
   <media:community>
    <media:starRating count="25" average="4.84" min="1" max="5"/>
    <media:statistics views="6666"/>
   </media:community>
  </media:group>
 </entry>
 <entry>
  <id>yt:video:B4m2RVFLbME</id>
  <yt:videoId>B4m2RVFLbME</yt:videoId>
  <yt:channelId>UCzd1PDbjInGwU5mLLea_Cyw</yt:channelId>
  <title>Real-time face recognition with Deep Learning technology</title>
  <link rel="alternate" href="https://www.youtube.com/watch?v=B4m2RVFLbME"/>
  <author>
   <name>Vicomtech-IK4</name>
   <uri>https://www.youtube.com/channel/UCzd1PDbjInGwU5mLLea_Cyw</uri>
  </author>
  <published>2016-07-01T11:35:45+00:00</published>
  <updated>2017-07-11T03:54:43+00:00</updated>
  <media:group>
   <media:title>Real-time face recognition with Deep Learning technology</media:title>
   <media:content url="https://www.youtube.com/v/B4m2RVFLbME?version=3" type="application/x-shockwave-flash" width="640" height="390"/>
   <media:thumbnail url="https://i3.ytimg.com/vi/B4m2RVFLbME/hqdefault.jpg" width="480" height="360"/>
   <media:description>Real-time face recognition with Deep Learning technology

This video shows an automatic real-time video-surveillance system capable of recognizing people faces from challenging images, with remarkable differences from the images used for the registration, such as low quality ID card photos and online registrations of initially unknown people. This solution relies on the latest deep learning-based technologies for improving the face recognition rates.

More information:

Viulib® is a software library, a solution that collects, processes and analyzes real-time video images. Download Viulib: http://www.viulib.org/

Viulib is property of http://www.vicomtech.org/
Viulib Contact: viulib@vicomtech.org
General Contact: tech.transfer@vicomtech.org</media:description>
   <media:community>
    <media:starRating count="13" average="4.69" min="1" max="5"/>
    <media:statistics views="2714"/>
   </media:community>
  </media:group>
 </entry>
 <entry>
  <id>yt:video:QcCjmWwEUgg</id>
  <yt:videoId>QcCjmWwEUgg</yt:videoId>
  <yt:channelId>UCzd1PDbjInGwU5mLLea_Cyw</yt:channelId>
  <title>Real-time event detection for video surveillance applications</title>
  <link rel="alternate" href="https://www.youtube.com/watch?v=QcCjmWwEUgg"/>
  <author>
   <name>Vicomtech-IK4</name>
   <uri>https://www.youtube.com/channel/UCzd1PDbjInGwU5mLLea_Cyw</uri>
  </author>
  <published>2016-06-21T10:18:15+00:00</published>
  <updated>2017-08-01T11:58:03+00:00</updated>
  <media:group>
   <media:title>Real-time event detection for video surveillance applications</media:title>
   <media:content url="https://www.youtube.com/v/QcCjmWwEUgg?version=3" type="application/x-shockwave-flash" width="640" height="390"/>
   <media:thumbnail url="https://i2.ytimg.com/vi/QcCjmWwEUgg/hqdefault.jpg" width="480" height="360"/>
   <media:description>Events are detected in real-time in embedded platforms using optimized computer vision and machine learning algorithms. The detection process is based on the ability to detect and track individuals, and analyze their different trajectory and motion patterns. A calibration process helps to provide 3D context to the scene and therefore increasing the detection performance by reducing the number of False Positives.
Part of the developments were carried out in the framework of the EU FP7 project P-REACT (www.p-react.eu).
The developments are included in the Viulib libraries.

More information:

Viulib® is a software library, a solution that collects, processes and analyzes real-time video images. Download Viulib: http://www.viulib.org/
Viulib is property of http://www.vicomtech.org/
Viulib Contact: viulib@vicomtech.org
General Contact: its@vicomtech.org</media:description>
   <media:community>
    <media:starRating count="41" average="4.61" min="1" max="5"/>
    <media:statistics views="16977"/>
   </media:community>
  </media:group>
 </entry>
</feed>
