Masked Siamese

Masked Siamese Networks (MSNs) are a self-supervised learning approach that leverages Siamese architectures and masked image modeling to learn robust image representations. Current research focuses on improving MSN efficiency and effectiveness through variations in masking strategies (e.g., filling-based versus erase-based), incorporating multimodal data (like Electronic Health Records with medical images), and adapting the approach for different model architectures (including convolutional neural networks and vision transformers). This work is significant because it advances self-supervised learning, enabling improved performance on downstream tasks like image classification, object detection, and anomaly detection, particularly in scenarios with limited labeled data.