UnDEMoN: Unsupervised Deep Network for Depth and Ego-Motion Estimation

Keyphrases

• Six Degrees of Freedom 100%
• Depth Motion 100%
• Ego-motion Estimation 100%
• Camera Pose 100%
• Deep Network 100%
• Performance Comparison 50%
• Image Gradient 50%
• Superior Performance 50%
• Pose Estimation 50%
• Existing State 50%
• Monocular Image 50%
• Unsupervised Method 50%
• Driving Dataset 50%
• Supervised Methods 50%
• Depth Map 50%
• Bilinear 50%
• Spatial Reconstruction 50%
• KITTI 50%
• Unsupervised Deep Learning 50%
• Sampling Kernel 50%
• Pose Information 50%
• Coarse-to-fine Strategy 50%
• Depth Estimation 50%
• Monocular Stereo 50%
• Scale-aware 50%
• Absolute Scale 50%
• Dense Depth Map 50%
• Gradient Noise 50%
• Estimation Accuracy 50%
• Linear Sampling 50%
• Deep Learning Framework 50%
• Stereo Image Pair 50%
• Penalty Function 50%
• Reconstruction Loss 50%
• Temporal Reconstruction 50%

Computer Science

• Motion Estimation 100%
• Objective Function 100%
• Deep Learning 50%
• Performance Comparison 50%
• Image Gradient 50%
• Superior Performance 50%
• Pose Estimation 50%
• Learning Framework 50%
• Supervised Method 50%
• Depth Estimation 50%
• Estimation Accuracy 50%
• stereo-image 50%
• Unsupervised Method 50%
• Penalty Function 50%