Exploring the two-stage approach in neural network compression for object detection.
Xiao MengLixin YuZhiyong QinPublished in: ICMV (2018)
Keyphrases
- object detection
- neural network
- back propagation
- computer vision
- object class
- face detection
- neural network model
- artificial neural networks
- pedestrian detection
- compression ratio
- data compression
- neural network is trained
- image compression
- object classification
- deformable part models
- object categories
- neural nets
- background subtraction
- fuzzy logic
- human detection
- learning vector quantization
- image reconstruction from projections
- network architecture
- compression scheme
- activation function
- recurrent neural networks
- associative memory
- compression algorithm
- feed forward
- multi class
- pattern recognition
- data structure
- genetic algorithm
- multi layer
- network model
- training algorithm
- image classification
- knn
- moving objects
- feedforward neural networks
- compression rate
- image processing