Object Aware Contrastive Prior for Interactive Image Segmentation

Interactive Image Segmentation is a process of separating a user selected object from the background. This task requires building an effective class-agnostic segmentation model that performs well even on unseen categories. To achieve good accuracy with limited training dataset, it is important that...

Full description

Saved in:
Bibliographic Details
Main Authors: Mathur, Praful, Parwani, Shashi Kumar, Sen, Mrinmoy, Sheshadri, Roopa, Sharma, Aman
Format: Conference Proceeding
Language:English
Subjects:
Algorithms
and algorithms
Buildings
Computer vision
formulations
Image recognition and understanding
Image segmentation
Machine learning architectures
Performance evaluation
Self-supervised learning
Training
Training data
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Interactive Image Segmentation is a process of separating a user selected object from the background. This task requires building an effective class-agnostic segmentation model that performs well even on unseen categories. To achieve good accuracy with limited training dataset, it is important that the model has robust prior understanding of features of similar class objects. The model should also have good distinguishing capabilities of foreground objects with the background. In this paper, we propose Object Aware Click Embeddings (OACE) that represents user click aware foreground object features. OACE is obtained based on a prior network trained using the Contrastive Learning paradigm. The single-click object selection accuracy of our base interactive segmentation network is vastly improved with the OACE input. Additionally, we propose a Multi-Stage fusion approach to better utilize user click information. With the proposed method, we outperform existing state-of-the-art approaches by 21% on publicly available test-sets for click-based Interactive Image Segmentation.
ISSN:2642-9381
DOI:10.1109/WACV57701.2024.00063