activity
20192023
most citedNeighborhood Contrastive Learning for Novel Class Discovery

13 citations · 14 across the 3 of their papers we have counts for

collaborators
Showing cs.CVShow all

8 papers · 1 filter

cs.CV2023

Contrast, Stylize and Adapt: Unsupervised Contrastive Learning Framework for Domain Adaptive Semantic Segmentation

Tianyu Li, Subhankar Roy, Huayi Zhou +2

To overcome the domain gap between synthetic and real-world datasets, unsupervised domain adaptation methods have been proposed for semantic segmentation. Majority of the previous…

cs.CV20221 cited

Cooperative Self-Training for Multi-Target Adaptive Semantic Segmentation

Yangsong Zhang, Subhankar Roy, Hongtao Lu +2

In this work we address multi-target domain adaptation (MTDA) in semantic segmentation, which consists in adapting a single model from an annotated source dataset to multiple unann…

cs.CV202113 cited

Neighborhood Contrastive Learning for Novel Class Discovery

Zhun Zhong, Enrico Fini, Subhankar Roy +3

In this paper, we address Novel Class Discovery (NCD), the task of unveiling new classes in a set of unlabeled samples given a labeled dataset with known classes. We exploit the pe…

cs.CV2021

Curriculum Graph Co-Teaching for Multi-Target Domain Adaptation

Subhankar Roy, Evgeny Krivosheev, Zhun Zhong +2

In this paper we address multi-target domain adaptation (MTDA), where given one labeled source dataset and multiple unlabeled target datasets that differ in data distributions, the…

cs.CV2020

TriGAN: Image-to-Image Translation for Multi-Source Domain Adaptation

Subhankar Roy, Aliaksandr Siarohin, Enver Sangineto +2

Most domain adaptation methods consider the problem of transferring knowledge to the target domain from a single source dataset. However, in practical applications, we typically ha…

cs.CV2020

Motion-supervised Co-Part Segmentation

Aliaksandr Siarohin, Subhankar Roy, Stéphane Lathuilière +3

Recent co-part segmentation methods mostly operate in a supervised learning setting, which requires a large amount of annotated data for training. To overcome this limitation, we p…