Do we really need scene-specific pose encoders?

Keyphrases

• Encoder 100%
• Regressor 100%
• Pose Regression 100%
• Camera Pose 100%
• Regression Model 50%
• Proposed Architecture 50%
• Query Image 50%
• Visual Similarity 50%
• Model Estimates 50%
• Fully Connected 50%
• Multilayer Perceptron 50%
• Forward Pass 50%
• Fully Connected Layer 50%
• Competitive Results 50%
• Outdoor Localization 50%
• Visual Pose 50%
• Deep Convolutional Network 50%
• Model Training 50%
• Image Retrieval Framework 50%
• Shallow Architecture 50%

Computer Science

• Coder 100%
• Image Retrieval 50%
• Convolutional Neural Network 50%
• retrieval model 50%
• Multilayer Perceptron 50%
• Fully Connected Layer 50%