SG-NN: Sparse Generative Neural Networks for Self-Supervised Scene Completion of RGB-D Scans.
Angela DaiChristian DillerMatthias NießnerPublished in: CoRR (2019)
Keyphrases
- neural network
- camera tracking
- indoor scenes
- back propagation
- three dimensional
- video sequences
- laser scanner
- generative model
- depth images
- pattern recognition
- neural network model
- multi layer
- d scene
- artificial neural networks
- scene understanding
- fuzzy logic
- high dimensional
- recurrent neural networks
- optic flow field
- point cloud
- genetic algorithm
- nearest neighbor
- multilayer perceptron
- modular neural networks
- single image
- self organizing maps
- neural nets
- fault diagnosis
- feed forward neural networks
- augmented reality
- sparse representation
- indoor environments
- image sequences
- rgb d camera
- knn
- multi layer perceptron
- training algorithm
- dynamic scenes
- feed forward
- activation function
- hidden layer
- moving objects
- depth information
- bp neural network