CAE-GReaT: Convolutional-Auxiliary Efficient Graph Reasoning Transformer for Dense Image Predictions

Computer Science

• Vision Transformer 100%
• Convolutional Neural Network 80%
• local feature 60%
• Computational Cost 40%
• Computer Vision 20%
• Image Segmentation 20%
• Performance Gain 20%
• Instance Segmentation 20%
• Semantic Image 20%
• Network Architecture 20%
• Information Diffusion 20%