Image Descriptors and Keypoint Detection

We propose deep-learning solutions to extract image features and keypoint locations for applications such as image matching and 3D reconstruction. Opposed to hand-crafted features, such as SIFT, invariances are learned from example images.

Learned Local Features

• LIFT: Learned Invariant Feature Transform

Keypoint Detection

• TILDE: A Temporally Invariant Learned DEtector
• Learning to Assign Orientations to Feature Points

Floating-point Descriptors

• Deep Convolutional Descriptor Learning

Binary Descriptors

• Binary Descriptors
• BinBoost: Boosting Binary Keypoint Descriptors
• LDAHash: Binary Descriptor for Large-scale Applications
• BRIEF: Binary Descriptor for Low-cost Real-time Applications
• D-Brief: Efficient Discriminative Projections for Compact Binary Descriptors

Filter Learning

• Filter learning for image categorization and pixel classification
• Robust 3D Tracking with Descriptor Fields