Sharat Agarwal

My research interest lies in Computer Vision and Deep Learning; research topics include Active Learning, Data Fairness, and Domain Adaptation.

For comprehensive and effective training of deep models, our focus should be on proposing methods to utilize the available data efficiently. Thus, my research investigates visual data's contextual aspect using model's uncertainty to train deep networks effectively. I am interested in solving problems using less supervision.

Publications

Contextual Diversity for Active Learning

Contextual Diversity for Active Learning

Sharat Agarwal, Himanshu Arora, Saket Anand and Chetan Arora

ECCV20

PDFCODE
Does Data Repair Leads to Fair Models? Curating Contextually Fair Data to Reduce Model Bias

Does Data Repair Leads to Fair Models? Curating Contextually Fair Data to Reduce Model Bias

Sharat Agarwal, Sumanyu Muku, Chetan Arora, Saket Anand

WACV 2022

PDFCODE
Reducing Annotation Effort by Identifying and Labeling Contextually Diverse Classes for Semantic Segmentation Under Domain Shift

Reducing Annotation Effort by Identifying and Labeling Contextually Diverse Classes for Semantic Segmentation Under Domain Shift

Sharat Agarwal, Saket Anand, Chetan Arora.

WACV 2023

PDFCODE