BLOG
Pink glasses. Thoughts, research updates, and reflections from my journey through computer vision, multi-task learning, and the intersection of art and AI.
🎺🎺🎺 Graduated! 🎺🎺🎺
PhD graduation announcement with thesis on "Information Sharing Methods for Multi-Task Learning". The thesis examines information sharing for Multi-Task Learning (MTL) in multimedia and computer vision domains.
Read MoreTindART at ACM MM 2020
Visual arts recommender system using multi-task learning. A personal visual arts recommendation system that helps users discover artworks based on their preferences.
Read MoreOmniEyes: Analysis and synthesis of painted eyes at MMM 2020
Generative modeling of artistically painted eyes in portrait paintings. Exploring the patterns and techniques used in painting eyes across different artistic styles.
Read MoreArt, Color and Emotion at ACM MM 2019
Exploring emotion in art through color analysis. The ACE system enables users to navigate and examine emotional dimensions within art collections through color analysis.
Read MoreMany Task Learning With Task Routing at ICCV 2019
Method for routing data flows in multi-task learning models. Introduces the Task Routing Layer (TRL) that assigns network units to task-specific subsets, enabling 312 classification tasks in one model.
Read MoreFeaturewise Transformations in Multi-Task Learning at CIIT 2019
Invited talk on featurewise transformations in multi-task learning. Exploring how feature-level transformations can improve task-specific learning in MTL architectures.
Read MoreLearning Task Relatedness in Multi-Task Learning at ICMR 2019
Research on understanding task relationships in multi-task learning scenarios. Investigating optimal methods for knowledge transfer between simultaneous tasks.
Read MoreTask Routing in Multi-Task Learning
Deep dive into task routing mechanisms for multi-task learning architectures. Exploring structured approaches to data flow in neural networks.
Read MoreOmniArt in ACM TOMM
Introduction of the OmniArt benchmark dataset featuring annotated artwork data. A large-scale artistic benchmark for computer vision research.
Read More