Lunit Tech Blog

• PseudoEdgeNet: Nuclei Segmentation only with Point Annotations
• C-MIL: Continuation Multiple Instance Learning for Weakly Supervised Object Detection
• SRM : A Style-based Recalibration Module for Convolutional Neural Networks
• Weight Standardization
• Video Object Segmentation using Space-Time Memory Networks
• Towards Universal Object Detection by Domain Attention
• FickleNet:Weakly and Semi-supervised Semantic Image Segmentation using Stochastic Inference
• Image Inpainting using Deep Learning
• Revisiting Self-Supervised Visual Representation Learning
• StyleGAN: A Style-Based Generator Architecture for GANs