CVC participates in Deep Learning Barcelona as a key sponsor and contributes with several CVC researchers


CVC is participating in the 3rd Deep Learning Barcelona Symposium (DLBCN) as a key sponsor and is contributing with several CVC researchers. CVC is present on the organizing committee as well as on talks and posters. This DLBCN edition expands its categories by adding a life sciences area as it is key in the current artificial intelligence technological revolution. The symposium will feature deep learning world-class leaders in Barcelona on 22 and 23 December.


Barcelona is hosting on 22 and 23 December at UPC Campus Nord the third Deep Learning Barcelona Symposium (DLBCN). Deep learning is one of the engines of the current artificial intelligence technological revolution. This has been implemented in almost all audiovisual fields, such as speech recognition by virtual assistants or photo file organizers. The trend has leaped to the field of life sciences, with applications related to new drugs development or support for medical diagnosis. For this reason, this 2021 symposium is presenting for the first time a session on applications to genomics, nutrition, detection of COVID-19 from cough, or medical imaging.

Deep Learning Barcelona Symposium aims to highlight Barcelona-developed research, bringing together top-notch researchers who are either developing their research in the city right now or had pursued part of their academic career here. The symposium presents 32 talks and 41 posters from international universities such as Stanford, MIT, or Columbia, multinational companies as Meta, Amazon, NVidia and ElementAI, and local research centers like Computer Vision Center (CVC).

CVC is participating in Deep Learning Barcelona Symposium (DLBCN) in diverse ways. It is a sponsor together with Meta AI (Facebook), UOC eHealth Center and UPC School of Telecommunications. CVC researchers Dimosthenis Karatzas (UAB-CVC), Sergio Escalera (UB-CVC) and Petia Radeva (UB-CVC) are part of the Organization Committee along with some other local institutions and companies in the field of artificial intelligence. Various talks and posters are presented by CVC scientists:





The CVC at the Artificial Intelligence & Big Data Congress 2021

The AI & Big Data Congress, after its cancellation in 2020 due to the COVID-19 pandemic, is back in Barcelona in 2021. The 14th and 15th of September will be held in AXA Auditorium Barcelona the 7th edition of this leading conference on AI & Big Data in the business sector, organized by CIDAI, Centre of Innovation for Data Tech and Artificial Intelligence. The first day will deliver live plus streaming sessions.  Second day sessions will be fully online.

The conference will present trends and best practices of pioneering research centers and companies, new technological developments and their application and success stories explained in detail in the field of the IA and the Big Data. Therefore, the CVC will play a key role in this meeting, presenting some of its most relevants results of its research projects. We detail the CVC participation hereafter:

September 14th

Dr. Petia Radeva, “Can Deep Learning help to individuals to eat healthier?“. Onsite (Auditorium) and streaming

September 15th (only streaming)

Dr. Ernest Valveny, “Multimodal deep learning technologies to unlock document intelligence

Dr. Antonio López, “The Deep Learning Era: Data is the Driver in Autonomous Driving

Hosted by Dr. Carlos Sierra


All the CVC researchers interested in attending can make her/his registration here

This congress, except the keynote talks, will be held in Spanish




Large participation of CVC researchers in ICDAR2021 Conference

The 16th International Conference on Document Analysis and Recognition, (ICDAR 2021), will be held at Lausanne (Switzerland) with a hybrid format (onsite and online) from 5th  to 10th  September.

This year’s utter programme is composed of top-notch oral sessions, workshops and competitions, among others, in which a large number of CVC researchers are participating. It is worth noting that the main conference will take place from 8th to 10th September.  We detail CVC participation hereafter:

Organization committee:

  • Josep Lladós, Program Committee Chair
  • Alicia Fornés, Tutorials Chair

Oral Sessions:

Workshop’s papers:

  • GREC (Session 1: Handwritten Graphics. Sunday, September 5th, 11:20-11:30h)

Albert Suso, Pau Riba, Oriol Ramos and Josep Lladós. “A Self-Supervised Inverse Graphics Approach for Sketch Parametrization”

  • GREC (Session 3: Handwritten Text and Annotations. Sunday, September 5th, 16:10-16:20h)

Pau Torras, Mohamed Ali Souibgui, Jialuo Chen and Alicia Fornés. “A Transcription Is All You Need: Learning to Align through Attention”

  • GLESDO (Sunday, September 5th, 16:35-16:55h)

Sanket Biswas, Pau Riba, Josep Lladós and Umapada Pal. “Graph-based Deep Generative Modelling for Document Layout Generation


  • Rubèn Tito, Minesh Mathew, C.V. Jawahar, Ernest Valveny and Dimosthenis Karatzas, “ICDAR 2021 Competition on Document Visual Question Answering”


  • Session 1 (Thursday, September 9th, 8:30-10h). Track: NLP for document understanding

Rubèn Pérez Tito, Dimosthenis Karatzas, and Ernest Valveny. “Document Collection Visual Question Answering

  • Session 1 (Thursday, September 9th, 8:30-10h). Track: Document analysis for social good

Adrià Molina, Pau Riba, Lluis Gomez, Oriol Ramos-Terrades, and Josep Lladós. “Date Estimation in the Wild of Scanned Historical Photos: An Image Retrieval Approach”

  • Session 1 (Thursday, September 9th, 8:30-10h). Track: Indexing and retrieval of documents

Pau Riba, Adrià Molina Rodríguez, Lluis Gomez, Oriol Ramos-Terrades, and Josep Lladós. “Learning to Rank Words: Optimizing Ranking Metrics for Word Spotting”

  • Session 2 (Thursday, September 9th, 10:30-12h). Track: Document synthesis

Sanket Biswas, Pau Riba, Josep Lladós, and Umapada Pal. “DocSynth: A Layout Guided Approach for Controllable Document Image Synthesis”

Workshops (pre-conference September 5th -7th):

Sunday, September 5th CVC researcher Pau Riba will be the chair of the Keynote talk and session 1, and Alicia Fornes will be the chair of the General Discussion 2. In addition, two CVC papers will be presented in session 1, Handwritten Graphics and session 3, Handwritten Text and Annotations.

Sunday, September 5th (Half-Day)

CVC paper “Graph-based Deep Generative Modelling for Document Layout Generation” by Sanket Biswas, Pau Riba, Josep Lladós and Umapada Pal, will be presented at 16.35h.

Monday, September 6th (Half-Day)

CVC researchers Ruben Perez and Dimosthenis Karatzas participate as co-organizers.

Tuesday, September 7th

CVC researchers Alicia Fornés and Dimosthenis Karatzas participate as co-organizers.


In addition, several CVC researchers are taking part in the Doctoral Consortium, a satelite event to the ICDAR main conference, (Tuesday, September 7th, 14:00 – 17:30h) as mentors; Josep Lladós, Dimosthenis Karatzas and Oriol Ramos Terrades, and others as PhD participants; Sanket Biswas and Rubèn Pérez Tito.

Finally, it should be noted the participation of the CVC director Josep Lladós in the Editorial of this IJDAR special issue paper: “Advanced Topics in Document Analysis and Recognition”.



Three CVC papers selected for the ICCV2021 Conference

CVC is more than happy to announce that three of its papers have been selected for the ICCV2021 Conference, the premier international computer vision event.
Two of them were from Joost van de Weijer’s group and the other one from Sergio Escalera’s. Hereafter you can find them with a summary.

TransferI2I: Transfer Learning for Image-to-Image Translation from Small Datasets
Yaxing Wang, Hector Laria Mantecon, Joost van de Weijer, Laura Lopez-Fuentes, Bogdan Raducanu
Learning image-to-image (I2I) translation requires large datasets. In this paper, the authors propose transfer learning theory for I2I and show that they can now train high-quality systems on very small datasets.

Generalized Source-free Domain Adaptation
Shiqi Yang, Yaxing Wang, Joost van de Weijer, Luis Herranz, Shangling Jui
The authors propose a new method that can adapt to new target domains while keeping performance on the source domain without the need for any labels on the target domain.

DeePSD: Automatic Deep Skinning And Pose Space Deformation For 3D Garment Animation
Hugo Bertiche, Meysam Madadi, Emilio Tylson, Sergio Escalera
The authors present a novel solution to the garment animation problem through deep learning. Their contribution allows animating any template outfit with arbitrary topology and geometric complexity. When giving a piece (or set) of clothes, our approach links it to a “human body skeleton” thus, both clothing and person move in unison. Additionally, small deformations to the outfit for a realistic and consistent result are added.

CVC researchers at CVPR 2021

Several CVC researchers have attended this year’s Computer Vision and Pattern Recognition conference (CVCPR 2021), the world’s biggest gathering on Computer Vision Science. Like other events impacted by the COVID-19 pandemic, this year’s edition has taken place virtually from June 19th to June 25th.

CVC’s presence at CVPR 2021 has been the following:

Accepted Paper at CVPR (Main Conference):

Yang, L. Herranz, Y. Cheng, M. Mozerov, “Slimmable compressive autoencoders for practical neural image compression”, Proc. International Conference on Computer Vision and Pattern Recognition (CVPR21), June 2021 [arxiv][slides][​​video​​]

Accepted Papers at CVPR Workshops:

  1. Katakol, L. Herranz, F. Yang, M. Mrak, “DANICE: Domain adaptation without forgetting in neural image compression”, CVPR Workshop and Challenge on Learned Image Compression (CLIC 2021), June 2021 [arxiv][slides][​​video​​].
  2. Wang, L. Herranz, J. Van de Weijer, “Continual learning in cross-modal retrieval”, CVPR Workshop in Continual Learning in Computer Vision (CLVISION 2021), June 2021 [arxiv].
  3. Rafael E. Rivadeneira, Angel D. Sappa, et al. “Results and Awards of the Second Thermal Image Super-Resolution (TISR’2021) Challenge”, IEEE Workshop on Perception Beyond the Visible Spectrum (PBVS 2021), June 2021 [paper]
  4. Ozge Mercanoglu Sincan, Julio C. S. Jacques Junior, Sergio Escalera, et al. “ChaLearn LAP Large Scale Signer Independent Isolated Sign Language Recognition Challenge: Design, Results and Future Research”, June 2021 [paper]

Workshop Women in Computer Vision: Petia Radeva as mentor