Wise Contrastive Loss

Wise contrastive loss leverages contrastive learning techniques to improve various image processing tasks, primarily focusing on addressing challenges in image-to-image translation, segmentation, and domain adaptation. Current research emphasizes patch-wise contrastive losses, often integrated with diffusion models or transformer architectures, to enhance feature representation and mitigate issues like catastrophic forgetting and inconsistent ground truth data. This approach shows promise in improving the accuracy and efficiency of medical image analysis, style transfer, and other computer vision applications where robust feature learning is crucial.

Papers

March 15, 2023

Zero-Shot Contrastive Loss for Text-Guided Diffusion Image Style Transfer
Serin Yang, Hyunmin Hwang, Jong Chul Ye
Zero Shot Learning Pre Trained Diffusion Model Image Style Transfer Wise Contrastive Loss

March 10, 2023

Adaptive Supervised PatchNCE Loss for Learning H&E-to-IHC Stain Translation with Inconsistent Groundtruth Image Pairs
Fangda Li, Zhiqiang Hu, Wen Chen, Avinash Kak
Image Pair Virtual Immunohistochemistry Immunohistochemistry Staining Wise Contrastive Loss H&E to IHC Stain Translation

February 22, 2023

A Global and Patch-wise Contrastive Loss for Accurate Automated Exudate Detection
Wei Tang, Kangning Cui, Raymond H. Chan
World Event Segmentation Method Patch Prediction Wise Contrastive Loss Hard Exudate

November 18, 2022

Delving into Transformer for Incremental Semantic Segmentation
Zekai Xu, Mingyi Zhang, Jiayue Hou, Xing Gong, Chuan Wen, Chengjie Wang, Junge Zhang
Convolutional Neural Network Transformer Based Incremental Semantic Segmentation Class Incremental Semantic Wise Contrastive Loss

August 27, 2022

CLUDA : Contrastive Learning in Unsupervised Domain Adaptation for Semantic Segmentation
Midhun Vayyat, Jaswin Kasi, Anuraag Bhattacharya, Shuaib Ahmed, Rahul Tallamraju
Contrastive Learning Semantic Segmentation Domain Adaptation Unsupervised Domain Adaptation Contrastive Loss Wise Contrastive Loss

July 27, 2022

Contrastive Image Synthesis and Self-supervised Feature Adaptation for Cross-Modality Biomedical Image Segmentation
Xinrong Hu, Corey Wang, Yiyu Shi
Image Synthesis Image to Image Cross Modality Self Supervised Feature Wise Contrastive Loss

June 8, 2022

ConFUDA: Contrastive Fewshot Unsupervised Domain Adaptation for Medical Image Segmentation
Mingxuan Gu, Sulaiman Vesal, Mareike Thies, Zhaoya Pan, Fabian Wagner, Mirabela Rusu, Andreas Maier, Ronak Kosti
Contrastive Learning Domain Adaptation Medical Image Segmentation Unsupervised Domain Adaptation Contrastive Loss Wise Contrastive Loss

Wise Contrastive Loss

Papers

Zero-Shot Contrastive Loss for Text-Guided Diffusion Image Style Transfer

Adaptive Supervised PatchNCE Loss for Learning H&E-to-IHC Stain Translation with Inconsistent Groundtruth Image Pairs

A Global and Patch-wise Contrastive Loss for Accurate Automated Exudate Detection

Delving into Transformer for Incremental Semantic Segmentation

CLUDA : Contrastive Learning in Unsupervised Domain Adaptation for Semantic Segmentation

Contrastive Image Synthesis and Self-supervised Feature Adaptation for Cross-Modality Biomedical Image Segmentation

ConFUDA: Contrastive Fewshot Unsupervised Domain Adaptation for Medical Image Segmentation