Project description:

The goal of this project is to push for a breakthrough in language technologies with respect to visual signals, with a special focus on Sign Languages (SLs), in which visual signals carry the entire communicative burden. To do so, it will be based on novel interdisciplinary collaborations between specialists in computer vision and machine learning (to extract relevant features from visual communicative signals, discover patterns in the data, and quantitatively assess them for recognition and synthesis), linguists specialised in multimodal communication and sign language (to construct predictive linguistic theories based on low-level patterns found in the data), and specialists in computer graphics (to produce virtual characters with sign language and multimodal conversational abilities).

Main research field: Computer Graphics, Animation, Deep Learning

Tasks to be performed: Research and innovation in topics related to open tools for 3D virtual characters animation,



including machine learning techniques.

- Proposal of a Deep Learning approach for efficient handshape recognition

- Evaluation and improvement of the proposed DL-based approach

- Use the findings gathered in the above tasks to enhance GTI’s current tools and algorithms dedicated dedicated to Avatar Animation

Group and complement: Group 3 + Level ñ

Dedication and working hours: Part. time (30h/week)

Planned remuneration approx: 28.414,28€ gross per year

Financing fund: Proyecto PRESP08722 - AEI - MDM-CEX2021-001195-M - DTIC - LOZANO, Angel - Unidades de Excelencia María de Maeztu - Departamento de Tecnologías de la Información y las Comunicaciones, financiado por MICIU/AEI/10.13039/501100011033. Esta contratación está financiada por el Plan Estatal de Investigación Científica, Técnica y de Innovación 2021-2023.

📌 ENG-PSR--27 - Specialist support technician: Sign Language, Deep Learning and Avatar Animation (Barcelona)
🏢 Pompeu Fabra University
📍 Barcelona