CVC CVC

CVC people at NeurIPS 2018

by

We had a good representation of CVC people at this year’s NeurIPS which took place on the first week of December in Montreal. The articles presented were the following:

‘Memory Replay GANs: learning to generate images from new categories without forgetting‘, by authors Chenshen Wu, Luis Herranz, Xialei Liu, Yaxing Wang, Dr. Joost van de Weijer and Dr. Bogdan Raducanu (all from CVC);

‘Image-to-image translation for cross-domain disentanglement‘, by authors Dr. Abel González, Dr. Joost van de Weijer and Dr. Yoshua Bengio;

‘TADAM: Task dependent adaptive metric for improved few-shot learning‘, by authors Boris N. Oreshkin, Pau Rodriguez (CVC member) and Dr. Alexandre Lacoste

At the Metalearning Workshop:

Cross-Modulation Networks for Few-Shot Learning, by authors Hugo Prol (CVC master student), Vicent Dumoulin and Luis Herranz.

Dr. Antonio López invited speaker at the First Forum on Autonomous Vehicles

by

Dr. Antonio Lopez took part of the First Forum on Autonomous Vehicles, organized by the Association of Industrial Engineers of Catalonia, that was held the 27th of November at Barcelona.

The event was celebrated in order to reflect on the changes that may arise from the implementation of the autonomous vehicle. In this first edition, called “between myth and reality”, it was discussed the issue of the autonomous car not only from the technological current but also from other areas such as Law, Sociology or Philosophy.

Dr. Antonio López was an invited speaker of the technological part, in which he talked about Artificial Intelligence for Autonomous Driving and presented his current research and recent projects such as Carla or Synthia.

CVC at Smart City Expo 2018

by

CVC was present at this year’s Smart City Expo presenting its technology in Computer Vision to a high number of enterprises and companies who visited us in our stand. The Smart City Expo is a great opportunity to meet game changes all over the world, who gather in Barcelona for one week in order to exchange ideas and business opportunities. A day after the Expo, we had a visit of more than 50 delegates from over 10 countries invited by ACCIO in order to know our Research Center and explore possible sinergies and collaborations.

GIANA Sub-challenge on gastrointestinal image analysis awards, MICCAI 2018 ENDOVIS Challenge

by

GIANA 2018 Sub-challenge on Gastrointestinal Image Analysis took part on September 1 during MICCAI 2018 conference, held in Granada, Spain. GIANA sub-challenge was part of the MICCAI 2018 EndoVis challenge, which comprised 4 different categories this year: Gastrointestinal Image ANAlysis, Surgical Workflow Analysis in the SensorOR, Robotic Scene Segmentation and CATARACTS.

GIANA 2018 Sub-challenge was jointly organized by 4 different institutions from Spain and France: Computer Vision Center/Universitat Autònoma de Barcelona (CVC-UAB), ETIS laboratory at Ecole nationale supérieure de l’Electronique et de ses Applications (ETIS-ENSEA), Hospital Clinic de Barcelona (HCB) and Hôpital Saint-Antoine AP-HP (HSA). Main organizers of the sub-challenge were Jorge Bernal from CVC-UAB and Prof. Aymeric Histace from ETIS-ENSEA.

GIANA Sub-challenge is the continuation of previous challenges (ISBI 2015, MICCAI 2015 and MICCAI 2017) on gastrointestinal image analysis. This year the sub-challenge consisted of three main different tasks: polyp detection and localization in videocolonoscopy, polyp segmentation in colonoscopy images and lesion detection and localization in wireless capsule endoscopy images.

Regarding this year’s challenge we would like to highlight two important novelties. First, we completely redefined the tasks associated to wireless capsule endoscopy images and we proposed participants to tackle lesion detection (including lesion classification) and localization. To efficiently train and test the different methodologies, a completely new database for wireless capsule endoscopy lesions (CAD-CAP dataset) has been created. This dataset contains over 3500 images comprising the following categories: normal, inflammatory lesions and vascular lesions.  We also offered participants on colonoscopy-related tasks the possibility of having a first glimpse of their performance by using the online evaluation tool that has been developed at CVC-UAB.

Teams had access to annotated training data from May 23rd and testing data was released on July 20. Final submission data of the results was September 3rd. All information related to the challenge was released in GIANA official website https://giana.grand-challenge.org though data was only accessible after registering on the website and acceptance of the rules for participation.

Awards for the challenge were sponsored by Société Française d’Endoscopie Digestive (SFED), which provided with 12 different prizes for a total of 2000 €. The following categories were defined: 1) Polyp Detection, 2) Polyp Localization, 3) Polyp Segmentation (SD images), 4) Polyp Segmentation (HD images), 5) Wireless Capsule Endoscopy Lesion Detection, 6) Wireless Capsule Endoscopy Lesion Localization,7) Overall Participation award, which considered only the teams that took part in all 3 previous categories and 8) Special GIANA award for meritorious contribution.

Winners went as follows: 

In the Polyp Detection category, the first prize (250 €) was awarded to the KM team composed by Satoshi Kondo.

In the Polyp Localization category, the first prize (250 €) was awarded to the ODS AI team composed by Vladimir Iglovikov and Anton Dobreenkii.

In the Polyp Segmentation (SD images) category, the first prize (250 €) was for the Penguin AI team, composed by Hanbo Chen, He Zheng, Yang Li, Jianhua Yao and Junzhou Huang.

In the Polyp Segmentation (HD images) category, the first prize (250 €) was awarded to the OUS team composed by Hemin Ali Qadir, Younghak Shin and Ilangko Balasingham.

In the Wireless Capsule Endoscopy Lesion Detection category, the first prize (250€) was for the KM team, composed by Satoshi Kondo.

In the Wireless Capsule Endoscopy Lesion Localization category, the first prize (250€) was awarded to the KM team composed by Satoshi Kondo.

With respect to Overall Participation award, there was a tie for the first prize (125 €) which was awarded to both KM (composed by Satoshi Kondo) and Penguin AI (composed by Hanbo Chen, He Zheng, Yang Li, Jianhua Yao and Junzhou Huang) teams.

The second prize (100 €) in the Overall Participation Award category was awarded to FanVoyage team composed by Ahmed Mohammed Kedir, Mohib Ullah, Sule Yildirim, Ivar Farup Marius Pedersen and Øistein Hovde.

ODS AI team, composed by Vladimir Iglovikov and Anton Dobreenkii, was the winner of the third prize (50 €) in Overall Participation Award category.

Finally, FanVoyage team, composed by Ahmed Mohammed Kedir, Mohib Ullah, Sule Yildirim, Ivar Farup Marius Pedersen and Øistein Hovde, won the 1st prize (100 €) in Special GIANA award category.

We want to acknowledge the quality of the contributions from all the teams. Slides presented in the challenge detailing the results, along with details of each of the methodology, are already published in GIANA website. Additionally, we aim to publish journal publications summarizing the results of the challenge. We also want to deeply thank the main organizers of EndoVis challenge (Stefanie Speidel, Lena Maier-Hein and Danail Stoyanov) for their help and support during challenge preparation. We look forward to organizing future iterations of this sub-challenge in the following years.

ExperimentAI kickoff session

by

On the 7th of November we had the official kick off to the ExperimentAI project, an initiative promoted and coordinated by the Computer Vision Center and funded by the Spanish Foundation for Science and Technology (FECYT) that brings Artificial Intelligence and Computer Vision closer to society.

The session started with an inaugural address by Joana Barbany, General Director of Digital Society at the Government of Catalonia and councillor of Sant Cugat del Vallès. It continued with an introduction of the project and the aims and objectives of all following sessions.

The introduction was then followed by an open on how Artificial Intelligence and Computer Vision are impacting our daily life, their uses and applications. The panel was composed by Dr. Fernando Vilariño, Associate CVC director, Alexandra Canet, CVC Communications Manager, CVC PhD students Laura López and Marc Masana, and last but not least, Sigrid Vila, Bruno Moya and Oriol Baeza. The last three are High School students who have spent a research stay at CVC as part of their Research School Projects.  In addition, all members of the public could contribute and thus detect the different challenges that this technology must face in their different areas of expertise.

ExperimentAI is a recurrent activity that will take place on a weekly basis from the 7th of November until the 20th of Februrary 2019 at the Library Living Lab. We will be presenting, every Wednesday, new experiences, prototypes and activities to understand the impact of AI and Computer Vision in our future and current society.

All information, calendar and registrations here: ExperimentAI website

Dr. Antonio López and Meritxell Bassolas at Vallès Visió

by

Dr. Antonio López and Meritxell Bassolas, CVC Knowledge and Technology Manager, were interviewed at Àmbit B30 (Vallès Visió) to talk about Autonomous Driving. In the program, CVC technologies on this field such as Elektra, Synthia and Carla are explained. You can watch the video here (in Catalan): http://www.vallesvisio.cat/ambit-b30-15-10-2018/

3 CVC Papers accepted at this year’s NeurIPS

by

CVC researchers have three accepted papers at this year’s Neural Information Processing Systems (NIPS) Conference which will be taking place in Montreal from the 2nd to the 8th of December 2018. Papers are the following: ‘Memory Replay GANs: learning to generate images from new categories without forgetting‘, by authors Chenshen Wu, Luis Herranz, Xialei Liu, Yaxing Wang, Dr. Joost … Read more

ACMCV 2018

by

CVC celebrated this year’s Annual Catalan Meeting on Computer Vision (ACMCV2018) that took place on the 20th of September. This event is annually organized by the Master in Computer Vision coordinators of the Catalan universities UAB, UOC, UPC and UPF. The aims of this meeting are to strengthen the Catalan academic and industrial computer vision network, to disseminate the most relevant works and to allow students from the Master in Computer Vision meet with members of the Catalan CV community.

This year’s edition counted, as always, with the presentation of the final master thesis from the Master in Computer Vision, a poster session and two keynote lectures carried out by Guillem Alenyà, from the Institute of Robotics and Industrial Informatics (IRI), who talked about the project CLOTHILDE and Marco Pedersoli from the Superior Technology School of Montreal (ETS Montreal), whose talk was about the evolution of his research during the last ten years.

The program and more details about the event can all be found at its website: http://acmcv.cat/

 

 

CVC researchers at this year’s ECCV2018

by

Several CVC researchers attended this year’s European Conference on Computer Vision (ECCV) that took place in Munich, Germany, from the 8 to 14th of September. CVC presented 6 papers in total to the main conference, and several other papers at the Conference’s Workshops:

Dr. Lluís Gómez, Dr. Marçal Rusiñol and Andrés Mafla presented their paper ‘Single Shot Scene Text Retrieval‘, Yaxing Wang presented his ‘Transferring GANs: generating images from limited data‘, Pau Rodriguez presented ‘Attend and Rectify: a Gated Attention Mechanism for Fine-Grained Recovery‘ and Dr. Sergio Escalera presented two papers: ‘Folded Recurrent Neural Networks for Future Video Prediction‘ and ‘Deep Structure Inference Network for Facial Action Unit Recognition‘ along with Dr. Meysam Madadi. Dr. Antonio López and Felipe Codevilla presented their paper ‘On Offline Evaluation of Vision-based Driving Models‘

In other hand, Raúl Gómez presented a poster and an oral presentation of his work ‘Learning to Learn from Web Data‘ at the ECCV 1st Multimodal Learning and Applications Workshop, Dr. Antonio López presented a demo of the CARLA simulator and Dena Bazazian was one of the selected Phd students to organise the Women in Computer Vision ECCV 2018 Workshop, and also presented a poster at the Epic Workshop her poster ‘Soft-PHOC Descriptor for End-to-End Word Spotting in Egocentric Scene Images‘ .

Have a look at our ECCV2018 Twitter Moment.

Related articles: 

6 CVC Papers Accepted At This Year’s ECCV

Dena Bazazian, Organiser Of The Women In Computer Vision ECCV 2018 Workshop

 

Dr. Fernando Vilariño elected as the new ENOLL President

by

Associate CVC Director and UAB Professor has been elected as the new Chairman of the European Network of Living Labs (ENOLL) at the Network’s annual Conference in Geneva.   Dr. Vilariño takes on the lead of a European Network that is working towards inclusive, responsible and Open Science and Innovation. With more than 150 active … Read more