no code implementations • 5 Jul 2023 • Shuai Chen, Subhradeep Kayal, Marleen de Bruijne
The paradigm of self-supervision focuses on representation learning from raw data without the need of labor-consuming annotations, which is the main bottleneck of current data-driven methods.
1 code implementation • EACL 2021 • Subhradeep Kayal
The concept of unsupervised universal sentence encoders has gained traction recently, wherein pre-trained models generate effective task-agnostic fixed-dimensional representations for phrases, sentences and paragraphs.
no code implementations • 25 Dec 2020 • Minh Le, Subhradeep Kayal
The ability to detect edges is a fundamental attribute necessary to truly capture visual concepts.
1 code implementation • 26 Jun 2020 • Subhradeep Kayal, Shuai Chen, Marleen de Bruijne
Self-supervised learning has proven to be invaluable in making best use of all of the available data in biomedical image segmentation.
no code implementations • 24 Apr 2020 • Subhradeep Kayal, Florian Dubost, Harm A. W. M. Tiddens, Marleen de Bruijne
Data augmentation is of paramount importance in biomedical image processing tasks, characterized by inadequate amounts of labelled data, to best use all of the data that is present.
1 code implementation • ACL 2019 • Subhradeep Kayal, George Tsatsaronis
Distributed representation of words, or word embeddings, have motivated methods for calculating semantic representations of word sequences such as phrases, sentences and paragraphs.
no code implementations • WS 2017 • Subhradeep Kayal, Zubair Afzal, George Tsatsaronis, Sophia Katrenko, Pascal Coupet, Marius Doornenbal, Michelle Gregory
In this paper we present a solution for tagging funding bodies and grants in scientific articles using a combination of trained sequential learning models, namely conditional random fields (CRF), hidden markov models (HMM) and maximum entropy models (MaxEnt), on a benchmark set created in-house.
no code implementations • 15 Apr 2016 • Subhradeep Kayal
In this paper, a popular theoretical model with it's origins in statistical physics and social dynamics, known as the Deffuant-Weisbuch model, is applied to the image segmentation problem.