HI THERE I’m Laia Tarrés
a Computer Vision PhD student based in sunny Barcelona.

PhD candidate in Computer Architecture

Under the supervision of Prof. Jordi Torres (UPC-BSC) and Prof. Xavier Giró-i-Nieto (Amazon Research). My current research aims at developing systems that enable sign language users to have automatic ways of translating online content. I am currently leading the Speech2Signs project, formerly led by Amanda Duarte, that tackles the task of automatic speech to sign language translation and viceversa through Deep Learning.

I am also participating in the X-prize rainforest, on the Providence Plus team, lead by Ferran Marqués (UPC) and Antonio Torralba (MIT).

• Research Grant: PRE2020-094223 .
• Publication: My current research on Sign Language Understanding has lead to publications [1], [2].
• Mentored students:
• Topic detection from sign language videos, BsC thesis, Alvaro Budria.
• Sign language translation with pseudo-glosses, BsC thesis, Patricia Cabot.
• Towards video alignment across cameras with sign language 2D poses, BsC thesis, Andrea Iturralde.
• Exploring Visual Representations for Sign Language Translation, MsC thesis, Maram A. Mohamed.

Masters in Advanced Telecommunication Technologies

Having selected the Multimedia Track, the program provides an understanding in depth on audiovisual systems, including deep learning techniques applied to language and vision.

• Internship: British Broadcasting Corporation (BBC), R&D team. Under the supervision of Marta Mrak, Xavier Giró-i-Nieto and Marc Górriz. I researched generative models for image colourisation. The internship lead to my Final thesis and publication [3].
• Internship: Institut de Robòtica i Informàtica Industrial (IRI). Under the supervision of Francesc Moreno and Xavier Giró-i-Nieto.
• Final thesis: Researched GAN-based image colourisation with Feature Reconstruction Loss.

Undergaduate in Telecommunications Technologies and Services Engineering

The program provides a solid grounding in the fundamentals of ICT engineering and the specific skills pertaining to the major I selected: Audiovisual Systems. Where I acquired the knowledge to conceive, design, implement and operate products, systems and services in the field of audiovisual systems engineering, including the fields of acoustics, image, audio, video and multimedia environments.

• Internship: CIT-UPC. I worked on collecting a Cyberbullying detection in teenagers dataset, under the supervision of Asunción Moreno and Antonio Bonafonte. It led to publication [4].
• Internship: MINSAIT. Participated in the Support to Univeristies and Research project, as a datascientist apprentice.
• Final thesis: I worked on Skin Lesion Classification with Residual Neural Network Ensemble. Under the supervision of Verónica Vilaplana and Marc Combalia

UPC School

The postgraduate course is one of the best in its field. It provides a short yet intensive course on the basis on deep learning and focuses on: computer vision, language, video, reinforcement learning. The curriculum is constantly updated and the instructors are experts on their fields and at various stages of their careers.

• Curriculum planning, budgeting and scheduling of instructors.
• Providing support to both instructors and students.

DeepLearning Barcelona Symposium

The symposium aims at showcasing research on deep learning carried out by scientists in/from Barcelona. This edition will count with 1 keynoite, 41 speakers and 40 posters.

• Planning and organizing the program and venue.
• Providing support to presenters, sponsors and other organizers.

Courses

• Postgraduate Course on artificial Intelligence, UPC School (2021-present).
• Undergraduate course of Data Science and Engineering, UPC (2021-present).
• Master in Advanced Telecommunication Technologies, UPC (2021-present).
• Master in Big Data, Barcelona Technology School (2021-present).
• Windmill ITN workshop at CTTC (March 2022).
• Deep Learning conference for Insight Center at the Dublin City University (May 2022).

Selection of lessons imparted

• Introduction to Deep Learning: Backpropagation, Tensors, Interpretability, MLP, CNN, Monitoring, Distillation [code]. Lab instructor.
• Sequence Modeling: RNN, Attention [code], Transformers. Lab instructor.
• Generative modeling: self-supervised, GAN, VAE and Diffusion models. Main theory and lab instructor.
• Speech: Introduction to Speech, Speech Enhancement, ASR. Lab instructor.
• Recommender systems: GCN and Recommender systems. Lab instructor.